阅读说明：本技术 一种路面状态检测装置的噪声去除算法 (Noise removal algorithm of road surface state detection device ) 是由 王�华 全威 郭慕瑶 田进保 魏志超 满永兴 于 2021-06-03 设计创作，主要内容包括：本发明的一种路面状态检测装置的噪声去除算法,通过如下步骤实现：步骤一,在特定的波长下给发光二极管输入多个具有线性比例关系的工作电流I-(i),相应得到多个反射光强度P-(i)；步骤二、利用数理统计方法对步骤一的多个工作电流I-(i)与对应的多个反射光强度P-(i)作线性回归,得到P-(i)＝β·I-(i)+ε-(i)(i＝1,2,…,n),步骤三、采用最小二乘法计算回归系数β,步骤四、将步骤三的反射光强度系数β来计算当前反射率。与现有技术相比采用最小二乘法的计算方式将噪声去除,使算出的反射光强度系数是一真值,无噪声参杂,利用此反射光强度系数来计算反射率的时候就会使计算出来的反射率较为精准,从而使路面状态检测装置的检测数据较为精准,大大提高了行车安全。(The invention discloses a noise removal algorithm of a road surface state detection device, which is realized by the following steps: step one, inputting a plurality of working currents I with linear proportional relation to the light-emitting diode under specific wavelength i Correspondingly obtain a plurality of reflected light intensities P i (ii) a Step two, utilizing a mathematical statistical method to carry out comparison on a plurality of working currents I in the step one i Corresponding to a plurality of reflected light intensities P i Linear regression to obtain P i ＝β·I i +ε i (i ═ 1, 2, …, n), step three, calculate the regression coefficient β with the least square method, step four, calculate the current reflectance with the reflected light intensity coefficient β of step three. Compared with the prior art, the method adopts a least square method to remove noise, so that the calculated reflected light intensity coefficient is a true value and has no noise doping, the calculated reflectivity is more accurate when the reflected light intensity coefficient is used for calculating the reflectivity,therefore, the detection data of the road surface state detection device is accurate, and the driving safety is greatly improved.)

1. A noise removal algorithm of a road surface state detection device is characterized by comprising the following steps:

step one, inputting a plurality of working currents I with linear proportional relation to the light-emitting diode under specific wavelength_i(i-1, 2, …, n), the photodiode obtains a plurality of reflected light intensities P corresponding to the specific wavelength_i(i＝1，2，…，n)；

Step two, utilizing a mathematical statistical method to carry out comparison on a plurality of working currents I in the step one_iCorresponding to a plurality of reflected light intensities P_iLinear regression to obtain P_i＝β·I_i+ε_i(i-1, 2, …, n), wherein the slope β of the linear regression line is expressed as the reflected light intensity coefficient per unit current, ε_iTotal noise generated for ambient temperature at unit current;

step three, calculating a regression coefficient beta by adopting a least square method, wherein the calculation formula is as follows:

wherein the content of the first and second substances,

and step four, substituting the reflected light intensity coefficient beta in the step three into a relational expression of which the reflectivity is equal to the ratio of the reflected light intensity coefficient to the emergent light intensity coefficient to obtain the reflectivity.

Technical Field

The invention relates to the field of traffic road detection, in particular to a noise removal algorithm of a road surface state detection device.

Background

The present inventors have developed a detection device that uses the cooperation of a light emitting diode and a photodiode to realize a detection result, which is based on the principle that the light emitting diode emits emission light at a specific wavelength, the emission light is reflected to the photodiode through a cover on the road surface, the photodiode reflects the intensity of the reflection light by measuring the number of received photons of the reflection light at the specific wavelength, the intensity of the reflection light is calculated to obtain the reflectance of the specific wavelength to the road surface state, and finally different road surface states are distinguished by setting a threshold value according to the reflectance. However, in practical applications, the intensity of the emitted light is greatly influenced by the ambient temperature, so that the detection device using the light emitting diode and the photodiode in cooperation causes a problem of inaccurate detection accuracy due to the influence of the ambient temperature.

Accordingly, the present inventors have made extensive studies on the above problems, and as a result, the present invention has been made.

Disclosure of Invention

The invention aims to provide a noise removal algorithm of a road surface state detection device, which eliminates the influence of noise generated by environment temperature change on the detection of a light sensor.

To achieve the above object, the solution of the present invention is as follows: a noise removal algorithm of a road surface state detection device is realized by the following steps:

step one, inputting a plurality of working currents I with linear proportional relation to the light-emitting diode under specific wavelength_i(i-1, 2, …, n) and the photodiode obtains a plurality of reflected light intensities P corresponding to the specific wavelength_i(i＝1，2，…，n)；

Step two, utilizing a mathematical statistical method to carry out comparison on a plurality of working currents I in the step one_iCorresponding multiple reflected light intensities P_iLinear regression to obtain P_i＝β·I_i+ε_i(i-1, 2, …, n), wherein the slope β of the linear regression line is expressed as the reflected light intensity coefficient per unit current, ε_iTotal noise generated for ambient temperature at unit current;

step three, calculating a regression coefficient beta by adopting a least square method, wherein the calculation formula is as follows:

wherein the content of the first and second substances,

and step four, substituting the reflected light intensity coefficient beta in the step three into a relational expression of which the reflectivity is equal to the ratio of the reflected light intensity coefficient to the emergent light intensity coefficient to obtain the reflectivity.

After the technical scheme is adopted, the noise removing algorithm of the road surface state detection device removes noise by adopting a calculation mode of a least square method, so that the calculated reflection light intensity coefficient is a true value and has no noise, and when the reflection rate is calculated by utilizing the reflection light intensity coefficient, the calculated reflection rate is relatively accurate, so that the detection data of the road surface state detection device is relatively accurate, and the driving safety is greatly improved.

Detailed Description

In order to further explain the technical solution of the present invention, the present invention is explained in detail by the following specific examples.

A noise removal algorithm of a road surface state detection device is realized by the following steps:

step one, inputting a plurality of working currents I with linear proportional relation to the light-emitting diode under specific wavelength_i(I-1, 2, …, n) for the photodiode at the specific wavelength according to a plurality of operating currents I_iCorrespondingly obtaining a plurality of reflected light intensities P_i(i ═ 1, 2, …, n); that is, n groups of observed values of samples (I) of working current and reflected light intensity are obtained under specific wavelength₁，P₁)，(I₂，P₂)，…，(I_n，P_n)；

Step two, utilizing a mathematical statistical method to carry out comparison on a plurality of working currents I in the step one_iCorresponding multiple reflected light intensities P_iLinear regression is carried out to obtain the mathematical structural formula P_i＝β·I_i+ε_i(i-1, 2, …, n), wherein the slope β of the linear regression line is the reflected light intensity coefficient, ε, at a unit current_iThe total noise generated by the ambient temperature at the current unit current;

step three, calculating a regression coefficient beta by adopting a least square method, wherein the calculation formula is as follows:

wherein the content of the first and second substances,

and step four, substituting the reflected light intensity coefficient beta in the step three into a relational expression of which the reflectivity is equal to the ratio of the reflected light intensity coefficient to the emergent light intensity coefficient to obtain the reflectivity. The calculation method of the intensity coefficient of the emergent light in the invention adopts the known technology, and specifically, the intensity of the emergent light of the LED can be obtained according to the reflectivity r of the emergent light of the LED under the specific wavelength of the known white paperTherefore, when the reflectivity of light under the road surface state at that time is calculated, the ratio of the reflected light intensity system at that time to the fixed emergent light intensity can be obtained only by steps one to three.

According to the noise removal algorithm of the road surface state detection device, noise is removed in a minimum two-multiplication calculation mode, the calculated reflected light intensity coefficient is a true value, noise is not mixed, the calculated reflectivity is accurate when the reflectivity is calculated by using the reflected light intensity coefficient, the detection data of the road surface state detection device is accurate, and driving safety is greatly improved.

In the present invention, the total noise epsilon generated by the ambient temperature under the current unit current in the step one comprises the thermal noise epsilon inside the photoelectric element₁And dark noise epsilon due to dark current₂The thermal noise is generated along with the random thermal motion of a large number of electrons, so that the large number of random events have Gaussian distribution type probability density, the thermal noise has symmetry and has a value which is compensated, when the number of measurement times is large enough, the arithmetic mean value is close to 0, and therefore the variance can beDark noise is shot noise and is also the result of a large number of independent random events, so the dark noise probability distribution is also a gaussian distribution, and is expected to be written as μ, and the variance is written as μThe expectation of the total noise epsilon is E (epsilon) ═ E (epsilon)₁)+E(ε₂) Mu, the variance of the total noise isThus, according to the mathematical formula of the second step, the expectation and variance are obtained for the observed values of the reflected light intensity samples under the n working currents obtained each time, and the observed values are respectively E (P)_i)＝β·I_i+μ(i＝1，2，…，n)， Calculating the estimator using the least squares of step threeTo expect, i.e. that

From the calculations, the estimatorsThe data is repeatedly changed after the unbiased estimation of the beta, the estimated value of the beta is repeatedly worked out, the estimated quantity has no systematic trend of overestimation or underestimation, the average value tends to the beta and is basically not influenced by noise, the calculated beta is a true value, and the influence of temperature on experimental data is basically eliminated by utilizing the reflectivity calculated by the beta.

The above embodiments are not intended to limit the scope of the present invention, and any suitable changes or modifications made by one of ordinary skill in the art should be construed as not departing from the scope of the present invention.