CN104913830A - Radar level instrument signal processing method based on curve fitting - Google Patents

Abstract

The invention relates to a radar level instrument signal processing method based on curve fitting. The method includes the steps of calculating the slope of each point of an echo rising edge, judging whether the duration time of the rising edge meets the minimum keeping time or not, determining a certain rising edge as a target estimation position through the maximum slope, fitting a parabola through the least square method according to data of a 60% to 90% section of a crest of the rising edge at the target estimation position, calculating the peak value and corresponding time, and finally accurately positioning a level instrument. By means of the method, the purpose of positioning the multi-diameter multi-echo crest is effectively achieved, the influences of the measurement environment and a target object itself on an echo rear edge are eliminated, and the positioning accuracy is improved.

Description

A kind of radar levelmeter signal processing method based on curve

Technical field

The present invention relates to field of industrial measurement, specifically a kind of radar levelmeter signal processing method based on curve.

Background technology

Thing position is industrial important parameter.The level measurement instruments such as existing static pressure method, float-ball type, ultrasound wave, have respective features and application scope.Radar levelmeter, by advanced treatment technology, has unique advantage, demonstrates the performance of its brilliance especially, play more and more important effect in the industrial production under the harsh measurement conditions such as high temperature, high pressure, steam, corrosive medium are strong.

Industry thing position, liquid level gauge operating mode complexity, cause the problem of measuring accuracy under the measuring condition as chaff interference, guided wave bar coating, foam, steam and turbulent flow.Typical algorithm is detection peak, and the prerequisite of algorithm application is that echo waveform is enough good.In actual application, waveform multipath effect often has multiple echo, and such as, wall echo causes the peak value of more than, is exactly in addition badly to determine peak value.

Existing placement technology is by carrying out linear interpolation realization to echo, along the impact being subject to measurement environment and destination object itself after peak value and echo, linear interpolation is difficult to compound actual waveform characteristic, and this makes to be difficult to determine correct echo time of arrival, and precision improvement is limited in one's ability.

Summary of the invention

For the deficiencies in the prior art, curve is fused in radar levelmeter location algorithm by the present invention, proposes a kind of radar levelmeter signal processing method based on curve.

The technical scheme that the present invention is adopted for achieving the above object is: a kind of radar levelmeter signal processing method based on curve, comprises the following steps:

Step 1: using the reflection echo under aimless original state as Noise Background N;

Step 2: compared with Noise Background N by the reflection echo X of actual measurement and disappear mutually, disappeared echo S mutually;

Step 3: to the smoothing filtering of echo S that disappears mutually, with the small amplitude fluctuations of filtering, obtains filtered echo S _w, calculate echo S _win the slope k often put of each rising edge, determine rising edge trough S _wtnto crest S _wcninitial time t _snwith end time t _en, i.e. the duration T of rising edge _hn=t _en-t _snif, rising edge duration T _hnbe less than minimum hold time T _hmin, then this rising edge is invalid;

Step 4: in the effective situation of rising edge, using the rising edge at the maximum some place of slope as target state estimator position;

Step 5: according to the crest S of target state estimator position _wcnthe data in 60% to 90% interval, utilize least square fitting para-curve;

Step 6: obtain peak value by the Fitting Calculation, and the time t of correspondence;

Step 7: R=(t-t accurately located by level meter _c) c/2, wherein, R be the thing distance of positions from, c is the light velocity, t _cfor the clock initial correction time.

The present invention has following beneficial effect and advantage:

1. the present invention is directed to multipath effect in actual application and produce multiple echo, by the feature determination target echo of rising edge;

2. along the impact being subject to measurement environment and destination object itself after the present invention utilizes least square fitting para-curve to solve peak value and echo;

3. the present invention utilizes least square fitting para-curve to promote the positioning precision of level meter.

Accompanying drawing explanation

Fig. 1 is method flow diagram of the present invention;

Fig. 2 (a) shows radar levelmeter echo situation 1;

Fig. 2 (b) shows radar levelmeter echo situation 2;

Fig. 2 (c) shows radar levelmeter echo situation 3;

Fig. 3 is curve schematic diagram of the present invention.

Embodiment

Below in conjunction with drawings and Examples, the present invention is described in further detail.

As shown in Figure 1, be schematic flow sheet of the present invention.First calculate the slope that echo rising edge is often put, and judge whether the rising edge duration meets minimum hold time; Then determine that certain rising edge is as target state estimator position by maximum slope; Again by the data separate least square fitting para-curve in 60% to 90% interval of the rising edge crest of target state estimator position, calculate the time of peak value and correspondence, finally obtain level meter and accurately locate.Concrete steps are as follows:

Step 1: using the reflection echo under aimless original state as Noise Background N;

Step 2: compared with Noise Background N by the reflection echo X of actual measurement and disappear mutually, disappeared echo S mutually;

Step 3: to the smoothing filtering of echo S that disappears mutually, the small amplitude fluctuations of filtering, obtains filtered echo S _w, calculate echo S _win the slope k often put of the n-th rising edge, determine rising edge trough S _wtnto crest S _wcninitial time t _snwith end time t _en, i.e. the duration T of rising edge _hn=t _en-t _snif, rising edge duration T _hnbe less than minimum hold time T _hmin, by invalid for this rising edge;

Step 4: maximum slope k _maxbe the slope of the n-th rising edge point, then using the n-th rising edge as target state estimator position;

As shown in Fig. 2 (a), there are 3 rising edges in radar levelmeter echo, the duration of 3 rising edges is all greater than minimum hold time, maximum slope k _maxbe the slope of the 2nd rising edge point, then using the 2nd rising edge as target state estimator position;

As shown in Fig. 2 (b), there are 3 rising edges in radar levelmeter echo, and the duration of the 3rd rising edge is all less than minimum hold time, then the 3rd rising edge is invalid, maximum slope k _maxbe the slope of the 2nd rising edge point, then using the 2nd rising edge as target state estimator position;

As shown in Fig. 2 (c), there are 3 rising edges in radar levelmeter echo, and the duration of the 3rd rising edge is all less than minimum hold time, then the 3rd rising edge is invalid, and the slope of the 3rd rising edge point is maximum slope k _max, because the 3rd rising edge is invalid, after excluding invalid rising edge, maximum slope k _maxbe the slope of the 2nd rising edge point, then using the 2nd rising edge as target state estimator position;

Step 5: the n-th rising edge crest S _wcnthe data separate least square fitting para-curve in 60% to 90% interval;

Step 6: obtain peak value by the Fitting Calculation, and the time t of correspondence;

As shown in Figure 3, radar n-th rising edge point A is rising edge crest S with the value of some B _wcn60% and 90%, by the data separate least square fitting para-curve in this interval, obtain dotted line, dotted line fixed point C is the peak value that curve calculates, and the corresponding time is t;

Step 7: R=(t-t accurately located by level meter _c) c/2, wherein, R be the thing distance of positions from, c is the light velocity, t _cfor the clock initial correction time.

Claims (1)

1., based on a radar levelmeter signal processing method for curve, it is characterized in that: comprise the following steps:

Step 1: using the reflection echo under aimless original state as Noise Background N;

Step 2: compared with Noise Background N by the reflection echo X of actual measurement and disappear mutually, disappeared echo S mutually;

Step 3: to the smoothing filtering of echo S that disappears mutually, with the small amplitude fluctuations of filtering, obtains filtered echo S _w, calculate echo S _win the slope k often put of each rising edge, determine rising edge trough S _wtnto crest S _wcninitial time t _snwith end time t _en, i.e. the duration T of rising edge _hn=t _en-t _snif, rising edge duration T _hnbe less than minimum hold time T _hmin, then this rising edge is invalid;

Step 4: in the effective situation of rising edge, using the rising edge at the maximum some place of slope as target state estimator position;

Step 5: according to the crest S of target state estimator position _wcnthe data in 60% to 90% interval, utilize least square fitting para-curve;

Step 6: obtain peak value by the Fitting Calculation, and the time t of correspondence;

Step 7: R=(t-t accurately located by level meter _c) c/2, wherein, R be the thing distance of positions from, c is the light velocity, t _cfor the clock initial correction time.