CN109800739A

CN109800739A - Electric heating rotary kiln temperature-detecting device

Info

Publication number: CN109800739A
Application number: CN201910130283.4A
Authority: CN
Inventors: 钟世波
Original assignee: Sichuan Zhongtianzhu Mining Co Ltd
Current assignee: Sichuan Zhongtianzhu Mining Co Ltd
Priority date: 2019-02-21
Filing date: 2019-02-21
Publication date: 2019-05-24
Anticipated expiration: 2039-02-21
Also published as: CN109800739B

Abstract

The invention provides an electric heating rotary kiln temperature detection device for rotary kiln electric heating equipment that integrates both video detection and temperature induction. The device includes a power supply, a processor module, a temperature induction module, a video detection module, and an analog-to-digital conversion module. . The device comprehensively utilizes the advantages of both, and achieves a good automatic temperature control effect through a relatively simple temperature control mechanism.

Description

Electric heating rotary kiln temperature-detecting device

Technical field

The invention belongs to technical field of temperature control, and in particular to a kind of electric heating rotary kiln temperature-detecting device and control Method.

Background technique

Rotary kiln device has efficient heat-transfer capability and good mixed performance, the forging suitable for a variety of raw materials of industry The processes such as burning, volatilization, isolation, are widely used in industries such as metallurgy, chemical industry, cement, paper dress, environmental protection.As typical multiple Miscellaneous industrial equipment, Rotary Kiln Control method are the hot and difficult issues of research.Since rotary kiln has multivariable, non-linear and strong coupling Characteristics, one relatively slow period of development experience that rotary kiln automatically controls, until Intelligent Control Theory starts to melt such as close After entering complex industrial process control, successfully solves the problem of Rotary Kiln Control and is continually applied in production practices, A large amount of theoretical result also produces therewith.So the research about Rotary Kiln Control theory is mostly mainly to grind with intelligent control Study carefully direction, common are fuzzy, neural network, expert system and hybrid intelligent control etc..

The control target of rotary kiln is rationally to determine rotary system the characteristics of thermal decomposition in advance before entering kiln according to raw material Wind, coal, material, kiln speed and each portion's temperature of system, pressure and other parameters, handle the correlation of rotary kiln and preheater, cooler well, Stablize the thermal regulation of whole system, safeguard kliner coating, extend the operation cycle of zero defects, total efficiency, realizes high-quality, stable yields and low Consumption production, while also to save energy consumption and reduce the pernicious gas content in exhaust gas.Wherein temperature is the key that Rotary Kiln Control. Rotary kiln generally can be divided into three temperature band: three preheating zone, burning zone and insulation belt temperature are with respectively different effects.Object Material preheats in preheating zone first, subsequently into burning zone.Burning zone is the major part of rotary kiln, in this temperature band into Row oxidation --- reduction reaction.There is no the material sufficiently reacted further to react in burning zone in insulation belt.Burning zone temperature Degree directly affects the performance of rotary kiln and the quality of product, is an important technological parameter, therefore it is required that must assure that in work Skill requires within certain deviation range of temperature, and keeps stablizing as far as possible.It is exactly to require burning zone temperature from control effect analysis Steady-state error or steady-state error very little is not present in degree control, and strong antijamming capability can be soon extensive once be interfered Original state is arrived again.

However, at actually control scene, environmental condition is sufficiently complex although having there is a large amount of theoretical result, therefore, Control effect is not still very ideal, analyzes reason, the controlling difficulties of calcined by rotary kiln temperature mainly include the following:

(1) physical-chemical reaction process complexity, heat transfer process are complicated in rotary kiln, and service condition and operating condition change Greatly, such as kiln liner, the thickness of kliner coating, raw slurry flow, moisture, ingredient, fuel coal quality variation are frequent, and there are non-linear, big Inertia, it is difficult to establish accurate mathematical model.

(2) key process parameter calcination band temperature is difficult to measure, and fiber ratio color temperature measurer is installed before kiln and measures burning zone There is detection lag, and cloud of dust mist serious interference of being pollinated in temperature of charge.

(3) big with ambient enviroment contact area since rotary kiln volume is larger, it is easy to by the interference of external environment, There are more uncertain factors, to increase the difficulty for accurately controlling calcination temperature.

(4) up to the present, most of rotary kilns are all based on conventional PID control, but rotary kiln operating condition is changeable, PID controller tends not to obtain satisfied control effect.On the one hand, during actual use in the related transducer of specific position The precision of device is difficult to grasp degree of aging in real time, is unfavorable for retrofit；On the other hand, mostly in practical application is by skilled Operator obtains relatively satisfactory result by constantly modifying controller parameter.Many times are not only expended in this way, and And once environment, condition change, then must setting parameter again, otherwise will be unable to the control effect obtained.

Summary of the invention

In view of the above analysis, the main purpose of the present invention is to provide one kind, and the above-mentioned electricity of rotary kiln in the prior art to be overcome to add Many defects existing for the automatic temperature-adjusting control of hot equipment, for example, temperature control effect is bad or Control system architecture is huge Greatly, cost is excessively high and control flow and algorithm it is excessively complicated, therefore the invention proposes a kind of detection of comprehensive video and temperature The advantages of automatic temperature control apparatus and control method of the rotary kiln electric heating equipment of both degree inductions, both comprehensive utilizations, By relatively simple temperature control system framework and control method, good automatic temperature-adjusting control effect is realized.

The purpose of the present invention is what is be achieved through the following technical solutions.

Technical solution of the present invention is related to a kind of electric heating rotary kiln temperature-detecting device, which includes power supply, processing Device module, temperature sense module, video detection module, analog-to-digital conversion module, temperature sense module are used to incude the temperature in kiln, Its signal exports after signal processing module is handled, and send to analog-to-digital conversion module, then handled by processor module, and regards Frequency detection module simultaneously also detects rotary kiln, and detection image signal, which is also sent to processor module, to be handled, processor The kiln temperature data that module is obtained according to temperature sense module are carried out referring to the kiln temperature data that video detection module obtains Correction.

Further, the reference voltage of 3.3V is converted to constant current using amplifier by temperature sense module, works as electric current It will generate voltage drop when flowing through thermal resistance (Rt), then be amplified the weak pressure drop signal by amplifier, by amplified letter Number be sent into analog-to-digital conversion module.

Further, video detection module includes video acquisition processing module, the module include deformation machine learning module, Rotary kiln kiln hood image feature value seeks module, and extends image function foundation and processing module:

Deformation machine learning module, for carrying out video acquisition rotary kiln kiln hood image to obtain the heating in kiln cylinder body Before situation image, picture centre is initially set up to image edge direction plane deformation updating formula, wherein due to video detection Module camera lens is not exclusively parallel with imaging plane, therefore has the image deformation on the injustice line direction, i.e. generation strain image:

Wherein, (x, y) indicates the initial position of image, (x_c,y_c) be correction after position, r indicate the shape away from imaging center Displacement is from k₁And k₂For it is described from center to edge direction on deformation coefficient, | | R_area(x, y) | | it is the mould of definite integral parameter Value；

In the kiln hood side for the rotary kiln that video detection module faces, 1/8,1/16,1/32 mark that 3 length are r is set Ruler, one end of three scales are arranged in kiln hood side and are located at kiln cylinder body end and respectively along the disc of kiln hood and the tangent formation of kiln cylinder body On, the other end of each scale is located at outside kiln cylinder body in kiln hood side and respectively along kiln hood and the disc of the tangent formation of kiln cylinder body Radial direction extends outwardly, and three scales are spaced each other 120 ° of settings, passes through length in video detection module acquired image Image of the smallest scale after deformation, i.e. length in image deformation and the ratio of its physical length are calculated as initial value, to be based on The ratio of length and its physical length of the mode of meanshift algorithm to other two scale in image deformation carries out respectively Iteration, the result of iteration is respectively as k₁And k₂；

Rotary kiln kiln hood image feature value seeks module, right for generating color image I to image progress compressing and converting The black white image answered i.e. monochrome image is I ', and monochrome image gray value g is by color space linear expression are as follows:

G=α_rI_r+α_gI_g+α_bI_b

Wherein α_r>=0, α_g>=0, α_b>=0, α_r+α_g+α_b=1

α in formula_r, α_g, α_bFor optional parameters, I_r, I_g, I_bIt is the color channel values of image I；

Building such as minor function V:

In formula, x, y are pixel, g_x, g_yThe respectively single color gradation value of x and y two o'clock, δ_{X, y}Color is converted into for image I The European measurement of x when the model space, y pixel carries out monochrome image dimensionality reduction to above-mentioned function V using GAUSS sliding average Processing, obtains different monochrome images:

Establish function L (x, y, σ, ρ)=ρ I ' (x, y) G (x, y, σ)

In formula, x, y are monochrome image coordinate value, and σ is scale factor, and ρ is zoom factor, and monochrome image is I ' (x, y)；

Extend image function to establish and processing module, for building strain image to the elongated area extended outside kiln cylinder body It is vertical to extend image function f_c(L (x, y, σ, ρ)), wherein L (x, y, σ, ρ) is standardized as [0,1], extends image function are as follows:

Wherein, λ is to extend slope, and the autocorrelation matrix of each pixel is calculated using Harris's matrix:

Wherein x, y are pixel coordinate, and N is image size, then extend the characteristic response function of image function are as follows:

R (x, y, c)=detA (x, y, f_c)-k (traceA (x, y, f_c))²

Wherein, k be invariant and its be k₁And k₂Arithmetic mean of instantaneous value；

It is obtained using definite integral is cumulative:

Further, processor module is obtained according to the temperature data that temperature sense module obtains with video detection module It is corrected with the kiln temperature data that kiln hood image represents, temperature control, the processor module packet is carried out according to correction result It includes heat setup module and temperature model establishes module:

Heat setup module, for the initial temperature and environment temperature T in rotary kiln₁When identical, if rotary kiln is in t moment Temperature be T (t), heat be Q (t), then have:

Q (t)=Q₁(t)+Q₂(t)

In formula, Q₁(t) --- the heat that rotary kiln itself generates；

Q₂(t) --- the heat of transmission；

In formula, C is the thermal capacity of rotary kiln, and s is the corrected rear and temperature of kiln hood image that video detection module obtains The color difference ratio of kiln hood image when for initial temperature；

Then the heat of rotary kiln is expressed as:

Temperature model establishes module, for obtaining to formula progress Laplace transform:

Establish the temperature model of rotary kiln are as follows:

K=aR, T=CR are enabled, then is had:

Wherein, K is amplification coefficient, and T is time constant, and τ is lag time.

Technical solution of the present invention has the advantage that

Both the present invention temperature control mode compound by creatively proposition video detection and temperature sense, comprehensively utilize Advantage, and specifically propose the concrete mode of image procossing in the circuit structure and video detection of temperature sense module, base It is corrected in the machine vision information of thermal imaging, gives relatively reliable and accurate temperature controlled model, pass through The practical control at MATLAB l-G simulation test and scene realizes good automatic temperature-adjusting control effect by verifying.

Detailed description of the invention

Attached drawing 1 is the structure principle chart of control device of the present invention；

Attached drawing 2 is the circuit diagram of temperature sense module of the present invention.

Specific embodiment

Be the structure principle chart of electric heating rotary kiln temperature-detecting device of the present invention referring to Fig. 1, the device include power supply, Processor module, temperature sense module, video detection module, analog-to-digital conversion module, temperature sense module is for incuding in kiln Temperature, signal are exported after signal processing module is handled, are sent to analog-to-digital conversion module, then carried out by processor module Reason, and video detection module simultaneously also detects rotary kiln, detection image signal also send to processor module and is handled, The kiln temperature number that the kiln temperature data that processor module is obtained according to temperature sense module are obtained referring to video detection module According to being corrected.

Preferably, the circuit diagram of temperature sense module as shown in Figure 2, temperature sense module is using amplifier by 3.3V's Reference voltage is converted to constant current, voltage drop will be generated when electric current flows through thermal resistance (Rt), then should by amplifier Amplified signal is being sent into analog-to-digital conversion module by weak pressure drop signal amplification.

Preferably, video detection module includes video acquisition processing module, which includes deformation machine learning module, rotation Rotary kiln kiln hood image feature value seeks module, and extends image function foundation and processing module:

Wherein, (x, y) indicates the initial position of image, (x_c, yc) be correction after position, r indicate away from imaging center Deformation distance, k₁And k₂For it is described from center to edge direction on deformation coefficient, | | R_area(x, y) | | for definite integral parameter Modulus value；

G=α_rI_r+α_gI_g+α_bI_b

Wherein α_r>=0, α_g>=0, α_b>=0, α_r+α_g+α_b=1

Building such as minor function V:

Establish function L (x, y, σ, ρ)=ρ I ' (x, y) G (x, y, σ)

R (x, y, c)=detA (x, y, f_c)-k (traceA (x, y, f_c))²

It is obtained using definite integral is cumulative:

Preferably, processor module according to the temperature data that temperature sense module obtains with video detection module obtain with The kiln temperature data that kiln hood image represents are corrected, and carry out temperature control according to correction result, which includes Heat setup module and temperature model establish module:

Q (t)=Q₁(t)+Q₂(t)

In formula, Q₁(t) --- the heat that rotary kiln itself generates；

Q₂(t) --- the heat of transmission；

Then the heat of rotary kiln is expressed as:

Establish the temperature model of rotary kiln are as follows:

K=aR, T=CR are enabled, then is had:

After establishing model, processor is according to the temperature automatic control of the model realization rotary kiln.

The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.

Claims

1. an electric heating rotary kiln temperature detection device, is characterized in that, this device comprises power supply, processor module, temperature sensing module, video detection module, analog-digital conversion module, and temperature sensing module is used for inducting the temperature in the kiln, its After the signal output is processed by the signal processing module, it is sent to the analog-to-digital conversion module, and then processed by the processor module, while the video detection module also detects the rotary kiln, and the detected image signal is also sent to the processor module for processing. The module calibrates the kiln temperature data obtained by the temperature sensing module with reference to the kiln temperature data obtained by the video detection module.

2. The electric heating rotary kiln temperature detection device according to claim 1, wherein the temperature sensing module converts the reference voltage of 3.3V into a constant current using an amplifier, and a voltage is generated when the current flows through the thermal resistance (Rt). Then, the weak voltage drop signal is amplified by the amplifier, and the amplified signal is sent to the analog-to-digital conversion module.

3. The electric heating rotary kiln temperature detection device according to claim 1, wherein the video detection module comprises a video acquisition and processing module, the module comprises a deformation machine learning module, a rotary kiln kiln head image feature value obtaining module, and Extended image function establishment and processing module:

The deformation machine learning module is used to first establish the plane deformation correction formula from the image center to the image edge direction before collecting the image of the kiln head of the rotary kiln to obtain the image of the heating condition in the kiln barrel. The plane is parallel, so there will be image deformation in the non-parallel direction, that is, a deformed image:

Among them, (x, y) represents the initial position of the image, (x _c , y _c ) is the corrected position, r represents the deformation distance from the imaging center, and k ₁ and k ₂ are the direction from the center to the edge. Deformation coefficient, R _area (x, y) is the modulus value of the definite integral parameter;

On the kiln head side of the rotary kiln facing the video detection module, three rulers of 1/8, 1/16, and 1/32 of length r are set, and one end of the three rulers is set on the kiln head side at the end of the kiln cylinder. And respectively along the circular surface formed by the tangent between the kiln head and the kiln cylinder, the other end of each scale is located outside the kiln cylinder on the side of the kiln head and is respectively along the radial direction of the circular surface formed by the tangent between the kiln head and the kiln cylinder. Outward extension, the three scales are set at 120° intervals from each other, and the deformed image of the scale with the smallest length in the image collected by the video detection module, that is, the ratio of the length in the image deformation to its actual length is calculated as the initial value, based on meanshift The algorithm is used to iterate the ratios of the lengths of the remaining two scales in the image deformation to their actual lengths, and the iteration results are respectively taken as k ₁ and k ₂ ;

Rotary kiln kiln head image feature value obtaining module is used to compress and convert the image to generate a color image I, the corresponding black and white image, that is, the monochrome image, is I', and the gray value g of the monochrome image is linearly represented by the color space as:

g = α _r I _r + α _g I _g + α _b I _b

where α _r ≥ 0, α _g ≥ 0, α _b ≥ 0, α _r + α _g + α _b =1

where α _r , α _g , α _b are optional parameters, and I _r , I _g , and I _b are the color channel values of the image I;

Build the following function V:

In the formula, x, y are pixel points, g _x , g _y are the monochrome grayscale values of x and y, respectively, δ _{x, y} is the Euclidean form of x, y pixels when image I is converted into color model space Measure, use GAUSS moving average to reduce the dimensionality of the monochromatic image to the above function V, and obtain different monochromatic images:

Establishing the function L(x, y, σ, ρ) = ρ · I′(x, y) · G(x, y, σ)

In the formula, x, y are the coordinate values of the monochrome image, σ is the scale factor, ρ is the scaling factor, and the monochrome image is I'(x, y):

The extension image function establishment and processing module is used to establish an extension image function f _c (L(x, y, σ, ρ)) for the extension area of the deformation image extending to the outside of the kiln cylinder, where L(x, y, σ, ρ) is normalized to [0, 1], and the extended image function is:

Among them, λ is the extension slope, and the Harris matrix is used to calculate the autocorrelation matrix of each pixel:

Where x, y are the pixel coordinates, and N is the image size, the characteristic response function of the extended image function is:

R(x, y, c)=detA(x, y, f _c )-k(traceA(x, y, f _c )) ²

where k is a constant factor and it is the arithmetic mean of _k1 and k2 _;

Using definite integral accumulation, we get:

4. The electric heating rotary kiln temperature detection device according to claim 1, wherein the processor module corrects the temperature data in the kiln represented by the kiln head image obtained by the video detection module according to the temperature data obtained by the temperature sensing module , and perform temperature control according to the calibration results. The processor module includes a heat setting module and a temperature model building module:

The heat setting module is used to set the temperature of the rotary kiln at time t as T(t) and the heat as Q(t) when the initial temperature of the rotary kiln is the same as the ambient temperature T1, then:

Q(t)=Q ₁ (t)+Q ₂ (t)

In the formula, Q ₁ (t)——the heat generated by the rotary kiln itself;

Q ₂ (t) - heat transfer;

In the formula, C is the heat capacity of the rotary kiln, and s is the color difference ratio of the kiln head image obtained by the video detection module after correction and the kiln head image when the temperature is the initial temperature;

Then the heat of the rotary kiln is expressed as:

The temperature model building module is used to Laplace transform the formula to get:

The temperature model of the rotary kiln is established as:

Let K=aR, T=CR, then we have:

Among them, K is the amplification factor, T is the time constant, and τ is the delay time.