CN105572051A - Polarized water turbidity measurement device, system and method - Google Patents

Abstract

The invention discloses a polarized water turbidity measuring device, measuring system and measuring method, which is characterized in that it comprises: a light source (1) for emitting a beam of monochromatic light I to a first lens (2), the The monochromatic light I sequentially passes through the first lens (2) and the first polarizer (5) and is split into incident light I _a and scattered light I _b at an angle of 90° by the beam splitter (8), The scattered light I _b enters the first photodetector (7), and the 90° surface scattered light I _a1 after the incident light I _a enters the measuring groove (3) passes through the second polarizer (6) and the second lens in sequence After (4), it enters the second photodetector (9); a temperature measuring sensor is arranged in the measuring tank (3). The invention provides a polarized water turbidity measurement device, measurement system and measurement method, which adopts the data processing mode of the least squares support vector machine to compensate the temperature, eliminate the measurement error, and improve the measurement accuracy.

Description

A polarized water turbidity measuring device, measuring system and measuring method

technical field

The invention relates to a polarized water turbidity measuring device, a measuring system and a measuring method, belonging to the technical field of liquid turbidity measurement.

Background technique

The turbidity of water is an optical effect in water samples, which refers to the degree of transparency of water samples caused by the scattering and absorption of light by suspended solid particles and impurities in water samples. The turbidity is related to the concentration of impurities and optical properties. In recent years, water pollution incidents have occurred from time to time, threatening the lives and health of the general public, and arousing the public's concern about water quality monitoring. Photoelectric measurement is widely used in the field of turbidity measurement due to its fast and accurate characteristics.

According to the international standard ISO7027 for turbidity measurement, the measurement of turbidity is to measure the scattered light in the direction of 90°. This measurement method has a good linear relationship in the low turbidity area, but due to the multiple scattering of scattered light in the high turbidity area It leads to measurement errors, and its measurement range is greatly limited, so it cannot be widely used in the detection of chemical turbidity. On the other hand, in actual measurement, background light and stray light in the incident liquid inevitably enter the detector, affecting the measurement accuracy of the entire turbidimeter.

In the disclosed Chinese invention (CN1087425), the inventor has adopted a single light source to enter the liquid surface, accepting 90° surface scattered light to be incident on the detector after being converged by the lens, although the inventor has adopted an automatic cleaning device to solve the problem caused by contamination. errors caused by damage. However, since this measurement method uses a detector, it cannot solve the measurement error caused by the shaking of the light source. At the same time, it selects non-polarized light for measurement, and the detector has no selectivity for various light sources, and cannot remove the influence of background light and multiple scattered light. In view of the above two problems, the turbidimeter measuring instrument designed by the patent of this invention eliminates most of the influence of background light and stray light interference by utilizing the characteristics of polarized light. At the same time, the design of dual optical paths can greatly eliminate the influence of light intensity jitter caused by the aging of the light source. Compared with the transmission method and the 90°scattering method, the optical path is shorter and the multiple scattering is less, so the measurement range is larger. A temperature sensor is also placed to measure the temperature inside the device. This device adopts the least squares support vector machine (LS-SVM) data regression processing method, and at the same time performs data fitting on temperature, scattered light intensity, original light intensity, and light intensity ratio. The combination method has better linearity on the problems of few samples, nonlinearity and high dimensionality, and solves the influence of temperature on turbidity, improving the accuracy.

Contents of the invention

The technical problem to be solved by the present invention is to provide a polarized water turbidity measurement device and measurement system with a single light source and dual optical path polarized light; further, the present invention provides a method for solving turbidity measurement caused by unstable light source error, and eliminate the influence of background light and multiple scattered stray light, and at the same time adopt the least squares support vector machine data processing method to compensate the temperature, eliminate the measurement error, and improve the measurement accuracy of the polarized water turbidity measurement method .

In order to solve the problems of the technologies described above, the technical solution adopted in the present invention is:

A polarized water turbidity measurement device, characterized in that it includes: a light source for emitting a beam of monochromatic light I to the first lens, and the monochromatic light I passes through the first lens and the second lens in sequence. After a polarizer, the light is split into incident light I _a and scattered light I _b at an angle of 90° by a spectroscopic mirror, and the scattered light I _b enters the first photodetector, and the incident light I _a enters the 90° behind the measuring groove. ° The surface scattered light I _a1 enters the second photodetector after passing through the second polarizer and the second lens in sequence; a temperature measuring sensor is arranged in the measuring groove.

The measuring tank comprises a U-shaped tank, and the U-shaped tank comprises two vertically arranged water inlet tanks and water outlet tanks, the bottom ends of the water inlet tanks and the water outlet tanks are connected by transverse grooves, and the surroundings of the water inlet tanks and the water outlet tanks are all An overflow tank is provided, and the bottoms of the two overflow tanks located on opposite sides of the water inlet tank and the water outlet tank are connected; the top of the water inlet tank is higher than the top of the water outlet tank; the inner wall of the water inlet tank is set There is at least one pair of baffles.

The included angle between the baffle plate and the inner wall of the water inlet tank is 45° to 75°.

The top of the water inlet groove is at least 2 cm higher than the top of the water outlet groove, and the cross-sectional area of the water inlet groove is larger than that of the water outlet groove.

The light source is 860nm near-infrared light; the spectroscope is a grating spectroscope; both the first photodetector and the second photodetector are phototransistors; the temperature measuring sensor is a DHT22 sensor.

A measurement system for polarized water turbidity, characterized in that: it includes the above-mentioned measuring device, a signal processing module, a display module and a baffle control circuit; the signal processing module includes sequentially connected amplifier circuits, filter circuits, AD conversion circuit; the amplifying circuit is respectively connected with the first photodetector and the second photodetector, and the display module is connected with the baffle control circuit; the signal processing module is connected with the display module.

A measuring method of a polarized water turbidity measuring device, characterized in that: comprising the following steps:

A system of linear regression equations using a least squares vector support machine:

y=W ^T Φ(x)+b(1)

$\mathrm{<span\; class="notranslate">\; min</span>}\mathrm{<span\; class="notranslate">\; J</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; W</span>}}^{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; C</span>}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

sty(x _k )=W ^T Φ(x _k )+b+e _k ,k=1,....n(3)

In formula (1), Φ(x) represents the influence function of each input water turbidity influencing factor on the turbidity value; W represents the function coefficient matrix of each water turbidity influencing factor; y represents water turbidity;

Equation (2) is the least squares vector fitting of Equation (1), e in Equation (2) represents the error; e _k represents the error of each water turbidity influencing factor; T represents the operation mode of the matrix; Adjustment factor for the proportion of control error; J(W,e) represents the constraint condition, choose the minimum value of W and e;

Formula (1), (2) is constrained to get formula (3), s.t. is the meaning of being constrained; formula (2) and formula (3) form a group of equations;

The Lagrangian equation is established by formula (3):

$\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; J</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; W</span>}}^{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

In formula (4), α represents the Lagrangian multiplier; α _i represents the Lagrangian multiplier of each water turbidity influencing factor; i=1～N;

According to the KKT condition, partial derivatives are obtained for W, e, α, b in formula (4):

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; W</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; e</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; Ce</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; W</span>}}^{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

Solve b and α from equations (5) to (8) to get the least squares support vector machine model:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$

In formula (9), f(x) represents water turbidity; k(x, x _i ) represents the radial basis function of each water turbidity influencing factor;

$\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; \{</span>\frac{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>{<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$

In formula (10), x represents the current measured value of each water turbidity influencing factor; x _i represents the previous training value of each water turbidity influencing factor; σ represents the width of the kernel function;

Put formula (10) into formula (9), get:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; \{</span>\frac{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>{<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 11</span>}<span\; class="notranslate">\; )</span>$

Bring each water turbidity influence factor into the formula (11) as input value, adjust the adjustment factor C and the kernel function width σ, the output value f(x) is the turbidity value; the output value f(x) and the standard turbidity liquid for calibration.

Described water turbidity influence factor comprises temperature, scattered light intensity, reference light intensity and light intensity ratio; Described scattered light intensity is the light intensity of scattered light _Ib , and described reference light intensity is the light intensity _of Ia1, and described The light intensity ratio is I _b /I _a1 .

The preparation method of described standard turbidity liquid is: the formazin solution of several different water turbidities is filled in described measuring tank successively, records its corresponding voltage value respectively, and described voltage value comprises I _b and I _a1 .

The water turbidity calculation method of the standard turbidity liquid is: calculate the light intensity ratio I _b /I _a1 according to the voltage value, and use the water turbidity y and the light intensity ratio x of the standard turbidity liquid as the horizontal and vertical coordinates respectively, linear Fitting to get a linear equation, y=k ₁ x+b ₁ , k ₁ and b ₁ are fixed values, the measured and calculated light intensity ratio is substituted into the y=k ₁ x+b ₁ equation to get the water turbidity the size of.

Compared with the prior art, the above technical scheme dedicated to the present invention has the following technical effects:

1. The measurement device and measurement system of the present invention proposes a turbidity measurement scheme with a single light source and two light paths to measure scattered light and reference light, which can effectively reduce the measurement error caused by the instability of the light source and the contamination of the detector .

2. The measuring device and measuring system of the present invention adopt polarizers to generate polarized light and analyzers to detect polarized light, which can greatly reduce the influence of background light and stray light and improve precision.

3. In the measuring device and measuring system of the present invention, the designed measuring tank utilizes the gravity difference to inject water, and the liquid passes through the baffle to ensure that the water flow is slow and the liquid level to be measured is level.

4. The measuring device and measuring system of the present invention improve the detection sensitivity of the system, simplify the entire measuring system, have a higher system integration degree, and have a more compact structure, and realize online rapid detection of water body turbidity under complex environmental conditions.

5. The present invention adopts the data processing method of the least squares support vector machine, compensates the temperature, eliminates the measurement error, and improves the measurement accuracy.

The present invention provides a polarized water turbidity measurement device, measurement system and measurement method, and proposes a turbidity measurement scheme with a single light source and dual detectors and a dual optical path to measure 90° scattered light and the light signal of the light source itself , effectively reducing the measurement error caused by the instability of the light source. Due to the use of polarized light detection, it can greatly reduce the error caused by background stray light and light that has been scattered into the detector by multiple scattering, with high precision and strong anti-interference ability. At the same time, compared with the traditional transmission method and 90°scattering method, using the principle of surface scattering has a shorter optical path and less multiple scattering, so the measurement range is larger. A temperature sensor is also placed to measure the temperature inside the device. This device adopts the least squares support vector machine (LS-SVM) data regression processing method, and at the same time performs data fitting on temperature, scattered light intensity, original light intensity, and light intensity ratio. The combination method has better linearity on the problems of few samples, nonlinearity and high dimensionality, and solves the influence of temperature on turbidity, improving the accuracy.

Instructions attached

Fig. 1 is a flowchart of the present invention;

Fig. 2 is a schematic diagram of the structure of the measuring tank in the present invention.

detailed description

The present invention will be further described below in conjunction with the accompanying drawings.

As shown in Figures 1 to 2, a polarized water turbidity measurement device is characterized in that it includes: a light source 1 for emitting a beam of monochromatic light I to the first lens 2, and the monochromatic light I After passing through the first lens 2 and the first polarizer 5 in turn, the light is split into incident light I _a and scattered light I _b at an angle of 90° by the beam splitter 8, and the scattered light I _b enters the first photodetector device 7, the 90° surface scattered light I _a1 after the incident light Ia enters the measurement groove 3 _passes through the second polarizer 6 and the second lens 4 in sequence and then enters the second photodetector 9; set in the measurement groove 3 There is a temperature measuring sensor.

Described measuring groove 3 comprises U-shaped groove, and described U-shaped groove comprises two vertically arranged water inlet grooves 12 and water outlet groove 10, and the bottom end of described water inlet groove 12 and water outlet groove 10 is communicated by transverse groove 13, and described water inlet groove 12 and the water outlet tank 10 are all provided with overflow tanks 14 around, and the bottoms of the two overflow tanks 14 located on the opposite sides of the water inlet tank 12 and the water outlet tank 10 are connected; the top of the water inlet tank 12 is higher than the At the top of the water outlet tank 10 ; at least one pair of baffles 11 are arranged on the inner wall of the water inlet tank 12 .

The overflow water flows into the recovery tank through the collection device for subsequent use.

The included angle between the baffle plate 11 and the inner wall of the water inlet tank 12 is 45° to 75°.

The top of the water inlet groove 12 is at least 2 cm higher than the top of the water outlet groove 10 , and the cross-sectional area of the water inlet groove 12 is larger than that of the water outlet groove 10 .

The light source 1 is near-infrared light of 860nm; the spectroscope 8 is a grating spectroscope; the first photodetector 7 and the second photodetector 9 are phototransistors; the temperature measuring sensor is a DHT22 sensor .

A measurement system for polarized water turbidity, characterized in that: it includes the above-mentioned measuring device, a signal processing module, a display module and a baffle control circuit; the signal processing module includes sequentially connected amplifier circuits, filter circuits, AD conversion circuit; the amplifying circuit is connected to the first photodetector 7 and the second photodetector 9 respectively, and the display module is connected to the baffle control circuit; the signal processing module is connected to the display module.

The light source 1 uses 860nm near-infrared light, which can minimize the influence of the chromaticity of the water sample on the turbidity measurement; the beam splitter 8 uses a grating type beam splitter to divide the monochromatic light emitted by the light source into a known light intensity ratio relationship For two beams of light with the same frequency, the grating-type beam splitter produces roughly the same color regions of the spectrum, and the resolution of the red light region is higher than that of the prism type; the photodetector is a phototransistor, which has high sensitivity and small size; the measurement tank 3 The water inlet is on the left, and the water outlet is on the right. The height difference between the two is used for water injection operation to reduce the water flow rate, avoid the generation of air bubbles, and maintain the level of the measured liquid level. The inner wall of the tank adopts the design of baffle plate 11. , so that the air bubbles in the water body can attach to avoid the influence of air bubbles on turbidity measurement.

The detailed path of the optical path of the present invention is as follows: the light beam I is produced by the light source 1, after passing through the first lens 2 and the polarizer 5 (i.e. the first polarizer), to be incident at 30° with the liquid surface to be measured, and generated by the beam splitter 8 The light beam I _a and the light beam I _b are separated, and the light beam I _b directly enters the first photodetector 7 . The light beam I _a is incident on the measurement groove 3, and the scattered light I _a1 is incident on the second polarizer 6, and then incident on the second photodetector 9 through the converging effect of the second lens 4. It is necessary to ensure that all the above-mentioned devices are on the same plane during measurement. , and the polarization direction of the polarizer is perpendicular to the plane of the incident direction of the light and the vibration direction.

The present invention utilizes the transmission method of light to measure the turbidity of the liquid to be measured, and its principle is as follows:

After passing through the polarizer, the light I emitted by the special light source of the present invention is attenuated to half of the original I′ according to the Malus law and then enters the beam splitter 8. I _a1 is the light of I _a passing through the beam splitter 8. I _b1 is the light that is scattered by the surface of I _b at 90° to the direction of the incident light. The quotient d of the two has a certain functional relationship with the turbidity of the current sample liquid to be tested, and the derivation is as follows:

$\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}}{{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}}$

Wherein: the voltage values of I _a1 and I _b1 after photoelectric conversion by the first photodetector and the second photodetector, I _b1 =I _b ;

According to Bill Lambert's law and mile scattering theorem, taking I _b1 = β ₁ I _b KT, I _a1 = β ₂ I _a into it can be obtained:

$\mathrm{<span\; class="notranslate">\; d</span>}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; b</span>}}\mathrm{<span\; class="notranslate">\; K</span>}\mathrm{<span\; class="notranslate">\; T</span>}}{{\mathrm{<span\; class="notranslate">\; \&beta;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; a</span>}}}$

Among them: β ₁ and β ₂ are spectral coefficients, K is related to the composition and properties of the liquid to be measured, T is the turbidity value of the liquid to be measured; β ₁ , β ₂ and K are all constants, and β _1* K/β ₂ = D to represent,

Available, $d=\frac{{\mathrm{\&beta;}}_1{I}_{b}KT}{{\mathrm{\&beta;}}_2{I}_b}=\frac{{\mathrm{\&beta;}}_{1}KT}{{\mathrm{\&beta;}}_2}$

Simplify the above formula to get, $d=\frac{{\mathrm{\&beta;}}_{1}KT}{{\mathrm{\&beta;}}_2}=\mathrm{D.}T$

It can be seen from the above formula that d is directly proportional to T, and the turbidity value T of the liquid to be tested can be obtained through the ratio d of I _a1 to I _b1 . Since the intensity of scattered light on the surface is small at low turbidity, it is difficult to be detected and the error is large. At the same time, at different temperatures, the Brownian motion of the liquid will be different, and the intensity of the scattered light will also be different, so the compensation calculation for the measurement temperature will be performed below.

A measuring method of a polarized water turbidity measuring device, characterized in that: comprising the following steps:

A system of linear regression equations using a least squares vector support machine:

y=W ^T Φ(x)+b(1)

$\mathrm{<span\; class="notranslate">\; min</span>}\mathrm{<span\; class="notranslate">\; J</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; W</span>}}^{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; C</span>}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; no</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

sty(x _k )=W ^T Φ(x _k )+b+e _k ,k=1,....n(3)

In formula (1), Φ(x) represents the influence function of each input water turbidity influencing factor on the turbidity value; W represents the function coefficient matrix of each water turbidity influencing factor; y represents water turbidity;

Equation (2) is the least squares vector fitting of Equation (1), e in Equation (2) represents the error; e _k represents the error of each water turbidity influencing factor; T represents the operation mode of the matrix; Adjustment factor for the proportion of control error; J(W,e) represents the constraint condition, choose the minimum value of W and e;

W=[W1, W2, W3, W4] is a set of four coefficients of the influence function of temperature, scattered light intensity, reference light intensity and light intensity ratio on turbidity, that is, a coefficient matrix;

Formula (1), (2) is constrained to get formula (3), s.t. is the abbreviation of subjectto, which means constrained; formula (2) and formula (3) form a group of equations;

Compared with the traditional linear regression method, formula (2) increases the sum of squares of the errors, and requires the minimum sum of squares of errors to prevent extreme values from dominating. At the same time, compared with the traditional method, the least squares support vector machine improves the linear programming problem to solve the problem of linear equations, thereby improving the calculation accuracy, and is sensitive to outliers. Since this problem is an unconstrained problem, it is difficult to handle. Establish the Lagrangian equation and turn it into a constrained problem.

The Lagrangian equation is established by formula (3):

$\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; J</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; e</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; W</span>}}^{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

In formula (4), α represents the Lagrangian multiplier; α _i represents the Lagrangian multiplier of each water turbidity influencing factor; i=1～N;

According to the KKT condition, partial derivatives are obtained for W, e, α, b in formula (4):

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; W</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 5</span>}<span\; class="notranslate">\; )</span>$

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; b</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 6</span>}<span\; class="notranslate">\; )</span>$

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; e</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; Ce</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 7</span>}<span\; class="notranslate">\; )</span>$

$\frac{<span\; class="notranslate">\; \&part;</span>\mathrm{<span\; class="notranslate">\; L</span>}}{<span\; class="notranslate">\; \&part;</span>{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; W</span>}}^{\mathrm{<span\; class="notranslate">\; T</span>}}\mathrm{<span\; class="notranslate">\; \&Phi;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; the\; y</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 8</span>}<span\; class="notranslate">\; )</span>$

Solve b and α from equations (5) to (8) to get the least squares support vector machine model:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 9</span>}<span\; class="notranslate">\; )</span>$

In formula (9), f(x) represents water turbidity; k(x, x _i ) represents the radial basis function of each water turbidity influencing factor; where k(x, x _i ) is a radial basis function, Linear kernel function, S-type kernel function, polynomial kernel function, generally take Gaussian kernel function;

$\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; \{</span>\frac{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>{<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 10</span>}<span\; class="notranslate">\; )</span>$

In formula (10), x represents the current measured value of each water turbidity influencing factor; x _i represents the previous training value of each water turbidity influencing factor; σ represents the width of the kernel function;

Put formula (10) into formula (9), get:

$\mathrm{<span\; class="notranslate">\; f</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}}{<span\; class="notranslate">\; \&Sigma;</span>}}{\mathrm{<span\; class="notranslate">\; \&alpha;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}\mathrm{<span\; class="notranslate">\; exp</span>}<span\; class="notranslate">\; \{</span>\frac{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; |</span>{<span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}{{\mathrm{<span\; class="notranslate">\; \&sigma;</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}}<span\; class="notranslate">\; \}</span><span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; b</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 11</span>}<span\; class="notranslate">\; )</span>$

Bring each water turbidity influence factor into the formula (11) as input value, adjust the adjustment factor C and the kernel function width σ, the output value f(x) is the turbidity value; the output value f(x) and the standard turbidity liquid for calibration.

Described water turbidity influence factor comprises temperature, scattered light intensity, reference light intensity and light intensity ratio; Described scattered light intensity is the light intensity of scattered light _Ib , and described reference light intensity is the light intensity _of Ia1, and described The light intensity ratio is I _b /I _a1 .

The preparation method of described standard turbidity liquid is: the formazin solution of several different water turbidities is filled in the described measuring tank 3 successively, respectively measure its corresponding voltage value, and described voltage value comprises I _b and I _a1 .

The water turbidity calculation method of the standard turbidity liquid is as follows: calculate the light intensity ratio I _b /I _a1 according to the voltage value, and use the water turbidity y and the light intensity ratio x of the standard turbidity liquid as the horizontal and vertical coordinates respectively, linear Fitting to get a linear equation, y=k ₁ x+b ₁ , k ₁ and b ₁ are fixed values, and substituting the measured and calculated light intensity ratio into the y=k ₁ x+b ₁ equation to get the water turbidity the size of.

Implementation steps of the present invention are as follows:

1. Open the water inlet valve and the water outlet valve, and slowly fill the measuring tank 3 with the liquid to be tested. Excess liquid overflows from the water inlet.

2. The control circuit sends an instruction to the photodetector, so that the photodetector is preheated before the measurement, and the measurement is performed after the temperature of the photodetector is stable.

3. After the light source outputs the corresponding frequency light beam through the polarizer, two beams of light are generated after passing through the beam splitter 8, and enter the measurement system. After the scattered light passes through the analyzer, the lens gathers the received light and transmits it to the photodetector , the photodetectors convert the received optical signals into electrical signals and transmit them to the signal processing module. The polarization analysis direction of the polarization analyzer can be adjusted to ensure the maximum light intensity obtained by the second photodetector.

4. The amplification circuit, filter circuit and AD conversion circuit in the signal processing module sequentially amplify, filter and AD convert the electrical signal, generate a digital signal and send it to the control circuit for processing, and the control circuit obtains the demodulated voltage value, And the ratio d is obtained from the formula, and the ratio d, the reference light intensity I _b , the scattered light intensity Ia ₁ , and the temperature value T are brought into the least squares support vector machine as the input variable Xi, and the obtained result y is processed by the MatLab toolbox After data training, b and α are obtained, after data fitting and calibration, they are finally sent to the display module for display.

The above is only a preferred embodiment of the present invention, it should be pointed out that for those of ordinary skill in the art, without departing from the principle of the present invention, some improvements and modifications can also be made. It should be regarded as the protection scope of the present invention.

Claims (10)

1. the measurement mechanism of a polarization type water turbidity, it is characterized in that, comprise: for launching the light source (1) of a monochromic beam I to the first lens (2), described monochromatic light I becomes incident light I in 90 ° angles with after the first polaroid (5) by spectroscope (8) light splitting through described first lens (2) successively _awith scattered light I _b, described scattered light I _benter the first photodetector (7), described incident light I _aenter 90 ° of surface scattering light I after measuring flume (3) _a1successively by entering the second photodetector (9) after the second polaroid (6) and the second lens (4); Described measuring flume is provided with sensor for measuring temperature in (3).

2. the measurement mechanism of a kind of polarization type water turbidity according to claim 1, it is characterized in that: described measuring flume (3) comprises U-type groove, described U-type groove comprises two vertically disposed intake chambers (12) and effluent trough (10), described intake chamber (12) is communicated with by transverse groove (13) with the bottom of effluent trough (10), the surrounding of described intake chamber (12) and effluent trough (10) is provided with overflow groove (14), be positioned at described intake chamber (12) and effluent trough (10) opposite side two described in overflow groove (14) bottom be connected, the top of described intake chamber (12) is higher than the top of described effluent trough (10), the inwall of described intake chamber (12) is provided with at least one pair of baffle (11).

3. the measurement mechanism of a kind of polarization type water turbidity according to claim 2, is characterized in that: described baffle (11) is 45 ~ 75 ° with the angle of intake chamber (12) inwall.

4. the measurement mechanism of a kind of polarization type water turbidity according to claim 2, it is characterized in that: the top of described intake chamber (12) exceeds at least 2 centimetres, the top of described effluent trough (10), the cross-sectional area of described intake chamber (12) is greater than the cross-sectional area of described effluent trough (10).

5. the measurement mechanism of a kind of polarization type water turbidity according to claim 1, is characterized in that: the near infrared light that described light source (1) is 860nm; Described spectroscope (8) is raster pattern spectroscope; Described first photodetector (7) and the second photodetector (9) are phototriode; Described sensor for measuring temperature is DHT22 sensor.

6. a measuring system for polarization type water turbidity, is characterized in that: comprise the measurement mechanism as described in any one of Claims 1 to 5, signal processing module, display module and baffle controls circuit; Described signal processing module comprises the amplifying circuit, filtering circuit, the A/D convertor circuit that connect successively; Described amplifying circuit is connected with described first photodetector (7) and the second photodetector (9) respectively, and described display module is connected with baffle controls circuit; Described signal processing module is connected with described display module.

7. the measuring method of the measurement mechanism of a kind of polarization type water turbidity according to any one of Claims 1 to 5, is characterized in that: comprise the following steps:

Utilize the equation of linear regression group of least square vector support machine:

y＝W ^TΦ(x)+b(1)

$\min J(W,e)=\frac{1}{2}{W}^{T}W+C\underset{k=1}{\overset{n}{\&Sigma;}}{e}_k^2---\left(2\right)$

s.t.y(x _k)＝W ^TΦ(x _k)+b+e _k,k＝1,....n(3)

In formula (1), Φ (x) represents the water turbidity factor of influence of each input to the influence function of turbidity value; W represents the function coefficients matrix of each water turbidity factor of influence; Y represents water turbidity;

Formula (2) carries out the matching of least square vector, the e Representative errors in formula (2) to formula (1); e _krepresent the error of each water turbidity factor of influence; T represents the compute mode of matrix; C representative is used for the regulatory factor of departure proportion; J (W, e) represents constraint condition, selects the minimum value of W and e;

Formula (1), (2) are carried out to about fasciculation and obtained formula (3), and s.t. is the affined meaning; Formula (2) and formula (3) composition system of equations;

Lagrange's equation is set up by formula (3):

$L(W,b,e,\mathrm{\&alpha;})=J(W,e)-\underset{i}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i\left(W^T\mathrm{\&Phi;}\right(x_i)+b-y(x_i\left)\right)---\left(4\right)$

In formula (4), α represents Lagrange multiplier; α _irepresent the Lagrange multiplier of each water turbidity factor of influence; I=1 ~ N;

According to KKT condition, to the W in formula (4), e, α, b ask local derviation to obtain:

$\frac{\&part;L}{\&part;W}=0\&RightArrow;\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i\mathrm{\&Phi;}\left(x_i\right)---\left(5\right)$

$\frac{\&part;L}{\&part;b}=0\&RightArrow;\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i=0---\left(6\right)$

$\frac{\&part;L}{\&part;e}=0\&RightArrow;{\mathrm{\&alpha;}}_i={\mathrm{Ce}}_k---\left(7\right)$

$\frac{\&part;L}{\&part;{\mathrm{\&alpha;}}_i}=0\&RightArrow;W^T\mathrm{\&Phi;}\left(x_i\right)+b+e_k-y\left(x_i\right)---\left(8\right)$

Solved b and α by formula (5) ~ formula (8), obtained least square method supporting vector machine model:

$f\left(x\right)=\underset{i=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_{i}k(x,x_i)+b---\left(9\right)$

In formula (9), f (x) represents water turbidity; K (x, x _i) represent the radial basis function of each water turbidity factor of influence;

$k(x,x_i)=\exp \left\{\frac{-\left|\right|x-x_i|{|}^2}{{\mathrm{\&sigma;}}^2}\right\}---\left(10\right)$

In formula (10), x represents the value when pre-test of each water turbidity factor of influence; x _ithe trained values of each water turbidity factor of influence before representative; σ represents kernel function width;

Bring formula (10) into formula (9):

$f\left(x\right)=\underset{j=1}{\overset{N}{\&Sigma;}}{\mathrm{\&alpha;}}_i\exp \left\{\frac{-\left|\right|x-x_i|{|}^2}{{\mathrm{\&sigma;}}^2}\right\}+b---\left(11\right)$

Bring in formula (11) using each water turbidity factor of influence as input value, regulate regulatory factor C and kernel function width cs, output valve f (x) is turbidity value; Output valve f (x) and standard turbidity solution are demarcated.

8. the measuring method of a kind of polarization type water turbidity according to claim 7, is characterized in that: described water turbidity factor of influence comprises temperature, scattered light intensity, reference light intensity and light intensity ratio; Described scattered light intensity is scattered light I _blight intensity, described with reference to light intensity be Ia ₁light intensity, described light intensity ratio is I _b/ I _a1.

9. the measuring method of a kind of polarization type water turbidity according to claim 7, it is characterized in that: the preparation method of described standard turbidity solution is: the formal hydrazine solution of several different water turbidity is filled with successively in described measuring flume (3), record its corresponding magnitude of voltage respectively, described magnitude of voltage comprises I _band I _a1.

10. the measuring method of a kind of polarization type water turbidity according to claim 8, is characterized in that: the water turbidity computing method of described standard turbidity solution are: calculate light intensity ratio I according to magnitude of voltage _b/ I _a1, using the water turbidity y of standard turbidity solution and light intensity ratio x as horizontal, ordinate, linear fit, obtains linear equation, y=k ₁x+b ₁, k ₁and b ₁for definite value, will measure and the light intensity ratio calculated substitution y=k ₁x+b ₁the size of water turbidity is just obtained in equation.