CN104197975A - Sensor measurement accuracy improving method based on measured value differential constraining - Google Patents

Abstract

The invention discloses a sensor measurement accuracy improving method based on measured value differential constraining. A user can determine the iterations number Nint to achieve dynamic adjustment of delta t in the iteration process and correct sensor output in a self-adaptation mode, the problem of delaying caused by arithmetic mean is solved, and the method has the self adaptability that the recursive mean does not have. The method is simple in operation, is applicable to real-time correction of a self information processing unit of an intelligent sensor, achieves the effect that software compensates the defect of hardware, overcomes the shortcoming of measurement inaccuracy caused by self resolution of the sensor to some extent, and improves measurement accuracy. The method can improve the sensor data measurement accuracy, further has certain noise removing capability and can be directly used for sensor output data processing.

Description

A kind of sensor measurement method for improving accuracy based on measured value differential constraint

Technical field

The invention belongs to sensor signal processing technology field, more specifically say, relate to a kind of sensor measurement method for improving accuracy based on measured value differential constraint, at the aspect such as Design of Smart Sensor and signal denoising, improve sensor measurement precision and remove noise.

Background technology

Along with the arrival of " information age ", as the means of obtaining information, sensor technology has obtained significant progress.Sensor application field is more and more extensive, and more and more higher to its requirement, demand is more and more urgent, and sensor technology has become one of important symbol of weighing a national science state-of-art.

Because sensor can change the signals such as various physical quantitys, chemical quantity and biomass into electric signal, make people can utilize computer realization measurement automatically, information processing and automatically control.But the output characteristics of sensor is subject to the impact of many environmental factors, as temperature, power-supply fluctuation, magnetic field etc., simultaneously sensor output is affected by self operational characteristic also, and this causes sensor output often to have certain error with actual value.In specific application scenario, such error can cause bad impact to whole system.

The error existing for sensor output and actual value, field application engineer also just for sensor image data (output) do some simple filtering and noise reductions, this respect also has the algorithm of a lot of maturations, if kalman filtering, arithmetic mean scheduling algorithm are all the algorithms that image data is done to post-processed, rarely have the method improving for sensor self measuring accuracy, most of method that improves sensor measurement precision is to have sensor manufacturing process level and explore new measuring method two aspects for the sensor of the particular type of certain applications.

Along with the development of microelectric technique and material science, sensor development and in application process, more and more combine with microprocessor, make sensor not only have vision, sense of touch, the sense of hearing, the sense of taste, also had the artificial intelligence such as storage, thinking and logic judgement, the development trend of sensor presents intellectuality.Intelligent sensor is exactly that a kind of sensor (passing through signal conditioning circuit) that has information detection and the information processing function concurrently is given intelligent combination with microprocessor.For the design of following intellectualized sensor, utilize the information process unit of sensor self can complete the processing of raw measurement data, thereby export measurement data more accurately.

The using method that can be applied at present raw measurement data processing has the methods such as arithmetic filtering, recurrence average filtering, weighted filtering, and these methods all exist some drawbacks.Although arithmetic filtering can be removed random disturbance, along with sampled point increase causes measurement data time delay large.Recursive Filtering and weighting Recursive Filtering, although each calculating moment can return measurement data, but filtered version is single, parameter adjustment blindly, does not have stronger adaptivity, therefore, wish such one simple and can dynamic adjusting method, can on this limited computing unit of sensor self information processing section, complete some simple sensor raw measurement data processing, utilize the method for software compensation hard ware measure defect, thereby improve the measuring accuracy of sensor self.

Summary of the invention

The object of the invention is to cause for sensor self resolution the problem that sensor measurement precision is not high, a kind of sensor measurement method for improving accuracy based on measured value differential constraint is provided, to a certain degree compensating hardware deficiency, according to measured value information, revise measured value, approach ideal value, to improve measuring accuracy.

For realizing above object, the present invention is based on the sensor measurement method for improving accuracy of measured value differential constraint, it is characterized in that, comprise the following steps:

(1), determine and calculate iterations N _int;

(2), obtain sensor raw measurement data, when sampling instant is N _intindividual sampling instant, starts trimming process, now, and current time i=N _int;

And make j=i-1, the initial generalized velocity of calculating sensor current time i (3):

$V_{\mathrm{ij}}^R=\frac{y_i-y_j}{(i-j)T_s};$

Wherein, T _sfor the sampling time of sensor, y _i, y _jfor sampling instant i, the measured value of j;

(4), make j=i-2;

(5), calculating sensor measured value generalized velocity:

$V_{\mathrm{ij}}=\frac{y_i-y_j}{(i-j)\mathrm{Ts}};$

(6) judgement:

If V _ij* V _{i (j-1)}< 0, the generalized velocity of current time i for: otherwise:

If a is V _ij>=0:

$V_{\mathrm{ij}}^R=\left\{\begin{array}{cc}{V}_{\mathrm{ij}};& V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j))\\ V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s):& V_{\mathrm{ij}}\&le;V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s\\ V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s);& V_{\mathrm{ij}}\&GreaterEqual;V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s)\end{array}\right.;$

If $V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j)),$ Make j=j-1, and return to step (5), otherwise, step (7) entered;

B), work as V _ijwhen <0

$V_{\mathrm{ij}}^R=\left\{\begin{array}{cc}{V}_{\mathrm{ij}};& V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j))\\ V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s):& V_{\mathrm{ij}}\&le;V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s\\ V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s);& V_{\mathrm{ij}}\&GreaterEqual;V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s)\end{array}\right.;$

If $V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j)),$ Make j=j-1, and return to step (5), otherwise, step (7) entered;

Wherein: $S_{\mathrm{Verr}}((i-j)*T_s)=\frac{M}{(i-j)*T_s};$

S _verr((i-j) * T _s) be generalized velocity error, be two sampled point time interval Δ t=(i-j) * T that gets _sfunction, the resolution that M is sensor;

(7), calculate generalized velocity as the generalized velocity after i sampling instant place sensor calibration, and according to formula:

$y_{i+2\_\mathrm{correct}}=y_i+V_{\mathrm{ij}}^R*T_s$

Obtain the correction output y of next moment i+1 _{i+1_correct}as the output valve of sensor;

(8), current time i adds 1, returns to step (3), asks for like this N _inteach sampling instant sensor real time correction output valve after individual sampling instant.

The object of the present invention is achieved like this.

This programme has proposed a kind of sensor measurement method for improving accuracy based on measured value differential constraint, determines iterations N _int, can realize and in iterative process, dynamically adjust Δ t, the output of adaptively correcting sensor, has overcome the adaptivity of utilizing time delay and the recurrence average etc. of arithmetic mean not to possess; The inventive method computing is simple, is applicable to intelligent sensor self information processing unit real time correction, realizes software compensation hardware deficiency, to a certain degree overcome that sensor self resolution causes measurement inaccuracy, improve measuring accuracy; The inventive method, not only can improve the accuracy of sensor measurement data, also has certain denoising ability, can be directly used in sensor output data processing.

Brief description of the drawings

Fig. 1 is the comparison diagram of linear transducer idealized characteristic and actual characteristic curve;

Fig. 2 inputs the actual output of lower sensor and the desirable curve map of exporting at random;

Fig. 3 is desirable output and actual output error curve in Fig. 2;

Fig. 4 is superimposed with output valve and graph of errors thereof after sine input lower sensor idea output, real output value and the correction of white noise;

Fig. 5 is a kind of embodiment process flow diagram of sensor measurement method for improving accuracy that the present invention is based on measured value differential constraint.

Embodiment

Below in conjunction with accompanying drawing, the specific embodiment of the present invention is described, so that those skilled in the art understands the present invention better.Requiring particular attention is that, in the following description, in the time that perhaps the detailed description of known function and design can desalinate main contents of the present invention, these are described in here and will be left in the basket.

First with the linear transducer input-output characteristic curve of Fig. 1, wherein M represents sensor intrinsic resolution, and Xres represents sensor input internal discrete sampling interval, and K is linear transducer ratio system.

In Fig. 1, strokes and dots straight line is ideal output characteristic, and owing to being subject to unavailable input-output characteristic so accurately in sensor process and digitized characteristic reality, dashed curve is discrete sampling input-output characteristic, is sensor actual measured property.Difference between the two shows the error that certainly exists between desirable actual value and measured value, and based on this, under given list entries, ideal value and measured value and error thereof are as shown in Figure 2.

In Fig. 2, for linear scale factor K=1, the sensor of intrinsic resolution M=0.1, under given random list entries, obtain desirable output and measure output and the two deviation curve thereof, wherein dotted line is desirable output, and solid line is that actual output is according to measured value information.As shown in Figure 3, observe and find that both deviations are less than intrinsic resolution M.Utilize like this thought of dynamic compensation, obtain revising measured value, approach ideal value, be specially:

1, the measurement value sensor generalized velocity in definition sensor discrete sampling situation, that is:

$\frac{\mathrm{\&Delta;y}}{\mathrm{\&Delta;t}}=\frac{y_i-y_j}{(i-j)T_s}---\left(1\right)$

Wherein y _i, y _jfor sampling instant i, the measured value of j, i ≠ j.

On this basis, definition measurement value sensor generalized velocity error is

Wherein M is sensor intrinsic resolution, the time interval that Δ t is twice measurement, and prove: in the instantaneous generalized velocity of moment i sensor and calculating generalized velocity, meet: V _r(i)≤V _c(i) ± S _verr(Δ t), wherein V _rand V (i) _c(i) be respectively the instantaneous generalized velocity of moment i and calculate generalized velocity, V _c(i) calculate V by (1) formula _r(i) in theory exist but the actual instantaneous generalized velocity that can not ask.

2, on self-defined generalized velocity basis, proposed to pass through iteration, constantly increase Δ t, realize and dynamically reduce generalized velocity error, make to calculate generalized velocity and approach gradually instantaneous generalized velocity, until iteration finishes, obtain and proofread and correct generalized velocity, according to proofreading and correct the anti-bearing calibration of separating measured value of generalized velocity, realize the measurement update to sensor, thereby improve measuring accuracy.

The lower calibration result of sinusoidal input as shown in Figure 4.Wherein dotted line is the ideal output that is superimposed with white noise, and dotted line is actual curve of output, and solid line is the curve after proofreading and correct.It is that 0 variance is 0.01 white Gaussian noise that the sinusoidal signal that is wherein input as 5+sin (t) is superimposed with average.The resolution M of sensor is 0.1.Output error relatively in, solid line is the graph of errors of timing not, dotted line is for proofreading and correct output and desirable output error curve, relatively find, after proofreading and correct, output error is starkly lower than not timing error, carries out many experiments simultaneously, and under ramp input, error mean and variance are as shown in table 1 below.

Table 1

(2) on linear transducer input-output characteristic basis, complete the theoretical proof of this bearing calibration validity, and provided after the method application, between the output valve after correction and idea output, statistical error average is wherein N _intfor setting iterations, actual iterations when i-j represents iteration termination.

As shown in Figure 5, it is consistent with summary of the invention for process flow diagram of the present invention, comprises iterations N is set _int, obtain sensor raw measurement data, get N _intindividual raw measurement data, just starts and proofreaies and correct.Iterations N is set _intdetermine as the case may be, as long as can ensure that iteration can stop.This two step this corresponding to step (1), (2).It should be noted that, sensor raw measurement data is ongoing, and it is also ongoing proofreading and correct, and in step (8), current time just can be updated to next moment like this.

Start after correction, adopt the generalized velocity after step (3)～(6) iterative computation sensor calibration, then obtain proofreading and correct output, this corresponding step (7) with the generalized velocity integration after this sensor calibration, final updating sampling instant, and return and calculate N _inteach sampling instant sensor real time correction output valve after individual sampling instant, proofreaies and correct output thereby realize, until sensor stops proofreading and correct.

Although above the illustrative embodiment of the present invention is described; so that those skilled in the art understand the present invention; but should be clear; the invention is not restricted to the scope of embodiment; to those skilled in the art; as long as various variations appended claim limit and definite the spirit and scope of the present invention in, these variations are apparent, all utilize innovation and creation that the present invention conceives all at the row of protection.

Claims (1)

1. the sensor measurement method for improving accuracy based on measured value differential constraint, is characterized in that, comprises the following steps:

(1), determine and calculate iterations N _int;

(2), obtain sensor raw measurement data, when sampling instant is N _intindividual sampling instant, starts trimming process, now, and current time i=N _int;

And make j=i-1, the initial generalized velocity of calculating sensor current time i (3):

$V_{\mathrm{ij}}^R=\frac{y_i-y_j}{(i-j)T_s};$

(4), make j=i-2;

(5), calculating sensor measured value generalized velocity:

$V_{\mathrm{ij}}=\frac{y_i-y_j}{(i-j)\mathrm{Ts}};$

(6) judgement:

If V _ij* V _{i (j-1)}<0, the generalized velocity of current time i for: otherwise:

If a is V _ij>=0:

$V_{\mathrm{ij}}^R=\left\{\begin{array}{cc}{V}_{\mathrm{ij}};& V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j))\\ V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s):& V_{\mathrm{ij}}\&le;V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s\\ V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s);& V_{\mathrm{ij}}\&GreaterEqual;V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s)\end{array}\right.;$

If $V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j)),$ Make j=j-1, and return to step (5), otherwise, step (7) entered;

B), work as V _ijwhen <0

$V_{\mathrm{ij}}^R=\left\{\begin{array}{cc}{V}_{\mathrm{ij}};& V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j))\\ V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s):& V_{\mathrm{ij}}\&le;V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s\\ V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s);& V_{\mathrm{ij}}\&GreaterEqual;V_{i(j-1)}^R+S_{\mathrm{Verr}}((i-j)*T_s)\end{array}\right.;$

If $V_{i(j-1)}^R-S_{\mathrm{Verr}}((i-j)*T_s)<V_{\mathrm{ij}}<V_{i(j-1)}^R+S_{\mathrm{Verr}}(T_s*(i-j)),$ Make j=j-1, and return to rapid (5), otherwise, step (7) entered;

Wherein: $S_{\mathrm{Verr}}((i-j)*T_s)=\frac{M}{(i-j)*T_s};$

S _verr((i-j) * T _s) be generalized velocity error, be two sampled point time interval Δ t=(i-j) * T that gets _sfunction, the resolution that M is sensor;

(7), calculate generalized velocity as the generalized velocity after i sampling instant place sensor calibration, and according to formula:

$y_{i+1\_\mathrm{correct}}=y_i+V_{\mathrm{ij}}^R*T_s$

Obtain the correction output y of next moment i+1 _{i+1_correct}as the output valve of sensor;

(8), current time i adds 1, returns to step (3), asks for like this N _inteach sampling instant sensor real time correction output valve after individual sampling instant.