阅读说明：本技术 一种新的电弧焊接的实时弧压跟踪方法 (Novel real-time arc voltage tracking method for arc welding ) 是由 雷艇 吴超群 顾世阳 胡家豪 于 2021-06-30 设计创作，主要内容包括：本发明属于焊接制造领域,具体涉及一种新的电弧焊接的实时弧压跟踪方法,包括：依托弧压模块以一定的频率采集直流脉冲TIG焊接的电弧电压；提取峰值弧压并S-G滤波,以滤波后均值作为直流脉冲TIG焊接的表征弧压；结合变异系数评价的统计学分析方法,求出弧压跟踪的上边界弧压和下边界弧压。进一步地,提出边界弧压跟踪策略,将实时弧压分别与边界弧压进行比较,在弧压模块内部设计一种二维三段模糊控制器；依据前述两种弧压比较结果,即弧压误差及其变化率,分别作为上段和下段弧压模糊控制器的输入(中段无需控制)；制定模糊控制规则表,确定弧压跟踪方向和步长,输入到运动控制单元,实现对焊接弧压实时调节。(The invention belongs to the field of welding manufacture, and particularly relates to a novel real-time arc voltage tracking method for arc welding, which comprises the following steps: collecting arc voltage of direct current pulse TIG welding at a certain frequency by depending on an arc voltage module; extracting peak arc voltage and filtering by S-G, and taking the average value after filtering as the representative arc voltage of direct current pulse TIG welding; and (4) calculating the upper boundary arc voltage and the lower boundary arc voltage of the arc voltage tracking by combining a statistical analysis method of coefficient of variation evaluation. Further, a boundary arc voltage tracking strategy is provided, real-time arc voltages are respectively compared with the boundary arc voltages, and a two-dimensional three-section fuzzy controller is designed in an arc voltage module; according to the two arc voltage comparison results, namely the arc voltage error and the change rate thereof, the two arc voltage comparison results are respectively used as the input of the upper-section arc voltage fuzzy controller and the lower-section arc voltage fuzzy controller (the middle section does not need to be controlled); and formulating a fuzzy control rule table, determining the arc voltage tracking direction and step length, inputting the arc voltage tracking direction and step length into the motion control unit, and realizing real-time regulation of the welding arc voltage.)

1. A novel real-time arc voltage tracking method for arc welding is characterized in that:

a calibration step: welding the circular seam of the tube plate, and collecting the arc voltage of direct current pulse TIG welding by an arc voltage module according to a set frequency; extracting peak arc voltage, carrying out S-G filtering, taking the average value after filtering as the representative arc voltage of direct current pulse TIG welding, and obtaining the upper boundary arc voltage and the lower boundary arc voltage of arc voltage tracking by adopting a statistical analysis method based on coefficient of variation evaluation on the representative arc voltage;

a real-time welding step: the arc voltage module collects arc voltage of direct current pulse TIG welding according to set frequency to obtain real-time arc voltage, the real-time arc voltage is input to a boundary arc voltage tracking model, the model respectively compares the real-time arc voltage with the boundary arc voltage in the calibration step, and the fuzzy controller selects different fuzzy control methods according to the numerical values of the real-time arc voltage and the boundary arc voltage and adjusts the arc voltage in real time after determining the arc voltage tracking direction and the step length.

2. The real-time arc voltage tracking method for electric arc welding of claim 1, wherein: in the calibration step, the filtered mean value is the filtered peak arc voltage mean value after extracting the peak arc voltage and carrying out S-G filtering, the filtered peak arc voltage mean value is calculated, the filtered arc voltage mean value is used as the characterization arc voltage of direct current pulse TIG welding, and the method for obtaining the upper boundary arc voltage and the lower boundary arc voltage comprises the following steps: selecting m calibration circular seams on the tube plate, solving the characterization arc pressure of the m circular seams, and performing statistical analysis on the characterization arc pressure of the m circular seams by adopting a coefficient of variation evaluation method, wherein the method specifically comprises the following steps: calculating the mean value of m characterization arc voltages, respectively subtracting the maximum value and the minimum value of the m characterization arc voltages from the mean value, taking the maximum difference value as a threshold value, respectively subtracting the sum and the sum of the mean value and the threshold value, and determining the arc voltage heelUpper boundary arc voltage V of trace_{Upper boundary}And lower boundary arc voltage V_{Lower boundary}。

3. The real-time arc voltage tracking method for electric arc welding of claim 1, wherein: the method for acquiring the real-time arc voltage comprises the following steps: receiving peak signals, delaying a set time T, collecting the average value of n peak arc voltages at intervals of the set time T, and averaging the n peak arc voltages to obtain the real-time arc voltage V at the peak stage_{Real time}The peak signal indicates that the welding current enters the pulse peak phase.

4. The real-time arc voltage tracking method for electric arc welding of claim 1, wherein: the real-time arc voltage tracking method specifically comprises the following steps:

after receiving the base value signal, V_{Real time}And V_{Upper boundary}And V_{Lower boundary}And comparing, calculating arc voltage errors and the change rate thereof, inputting the calculation result to a fuzzy controller, wherein the fuzzy controller comprises an upper arc voltage fuzzy controller and a lower arc voltage fuzzy controller, the difference value between the real-time arc voltage and the upper boundary arc voltage and the lower boundary arc voltage is the arc voltage error, the arc voltage error is divided by the set time t to be the arc voltage error change rate, and the base value signal represents that the welding current enters a pulse base value stage.

5. The real-time arc voltage tracking method for electric arc welding of claim 1, wherein: the fuzzy controller performs fuzzy control by using V_{Real time}And V_{Upper boundary}And V_{Lower boundary}And comparing, and performing selection execution according to a comparison result:

if the real-time arc voltage V_{Real time}Greater than or equal to upper boundary arc voltage V_{Upper boundary}Then the arc voltage error and the change rate of the arc voltage error and the change rate are used as the input of the upper arc voltage fuzzy controller;

if the real-time arc voltage V_{Real time}Lower boundary arc voltage V or less_{Lower boundary}Then the arc voltage error and the change rate of the two are used as the output of the lower arc voltage fuzzy controllerEntering;

if the real-time arc voltage V_{Real time}Greater than the lower boundary arc voltage V_{Lower boundary}And is less than the upper boundary arc voltage V_{Upper boundary}The trend of the welding gun does not need to be controlled.

6. The real-time arc voltage tracking method for electric arc welding of claim 1, wherein: the control method of the upper-section arc voltage fuzzy controller comprises the following steps:

step 6.1, defining the arc voltage error e of the upper arc voltage₁And its rate of change ec₁Membership function library u, control rule library R and clarification algorithm fd through respective quantization factor k_e1And k_ec1After zooming, inputting the data to an upper arc voltage fuzzy controller, namely two-dimensional input, and a fuzzification module D/F (digital/analog) outputs a quantized arc voltage error e₁And ec₁Respectively converted into fuzzy quantity E and EC, and the reasoning module M outputs fuzzy quantity U, and the sharpening module F/D converts the fuzzy quantity U into sharpening step length L₁Finally, the proportional factor k of the upper arc voltage_u1Zooming, calculating the tracking step length s of the upper arc voltage₁；

Wherein the quantization factor k_e1Is calculated by the step of calculating the arc voltage error e₁Has an actual value range of [ x_min，x_max]Error of arc voltage e₁The value range after fuzzification is [ y_min，y_max]Quantization factor k_e1The calculation formula is as follows: k is a radical of_e1＝(y_max－y_min)/(x_max－x_min) Similarly, the quantization factor k can be obtained_ec1Scale factor k_u1The calculation step is that the range of the feeding step length output by the upper-section arc voltage fuzzy controller is [ l_min，l_max]Feed step length s₁Has an actual value range of [ s ]_min，s_max]Then the scale factor k_u1The calculation formula is as follows: k is a radical of_u1＝(s_max－s_min)/(l_max－l_min)；

Step 6.2, calculating a sharpening step length L according to the upper-stage fuzzy controller₁By a scale factor k_u1After zooming, real is obtainedThe actual feed step s₁When the actuating mechanism adjusts the feeding of the welding gun, the welding gun is close to the tube plate, the arc length j between the welding gun and the tube plate is reduced, and the welding gun coordinate is changed into z_i+1。

7. The real-time arc voltage tracking method for electric arc welding of claim 1, wherein: the control method of the lower arc voltage fuzzy controller comprises the following steps:

step 7.1, defining the arc voltage error e of the lower segment arc voltage₂And its rate of change ec₂Membership function library u, control rule library R and clarification algorithm fd through respective quantization factor k_e2And k_ec2After zooming, inputting the data to a lower arc voltage fuzzy controller, namely two-dimensional input, and enabling a fuzzy module D/F to quantize the arc voltage error e₂And ec₂Respectively converted into fuzzy quantity E and EC, and the reasoning module M outputs fuzzy quantity U, and the sharpening module F/D converts the fuzzy quantity U into sharpening step length L₂Finally, the scale factor k of the lower segment arc voltage_u2Zooming, calculating the tracking step length s of the lower segment arc voltage₂；

Wherein the quantization factor k_e2Is calculated by the step of calculating the arc voltage error e₂Has an actual value range of [ x_min，x_max]Error of arc voltage e₂The value range after fuzzification is [ y_min，y_max]Quantization factor k_e2The calculation formula is as follows: k is a radical of_e1＝(y_max－y_min)/(x_max－x_min) Similarly, the quantization factor k can be obtained_ec2；

Scale factor k_u2The calculation step is that the value range of the output feeding step length of the lower arc pressure fuzzy controller is [ l ]_min，l_max]Feed step length s₂Has an actual value range of [ s ]_min，s_max]Then the scale factor k_u1The calculation formula is as follows: k is a radical of_u2＝(s_max－s_min)/(l_max－l_min)；

Step 7.2, calculating the sharpening step length L according to the lower fuzzy controller₂By a scale factor k_u2After scaling, the actual feeding step s is obtained₂At the moment, the actuating mechanism adjusts the feeding of the welding gun, the welding gun is far away from the tube plate, the arc length j between the welding gun and the tube plate is increased, and the welding gun coordinate is changed into z_i+1。

Technical Field

The invention relates to the field of welding manufacturing, in particular to an arc pressure control method for robot welding of a circular seam of a tube plate.

Background

The tube plate welding is a main processing mode for producing pressure vessels such as heat exchangers, steam generators, condensers and the like, and the service performance and the service life of the pressure vessel are directly determined by the welding quality of a tube plate joint. The tube plate welding is mainly direct current pulse TIG all-position welding, all welding seams of joints are annular welding seams, group seams are dense, heat input is concentrated, and thermal deformation influences the consistency of welding arc length between the annular seams and a tungsten electrode. An arc voltage tracking function is added in the robot arc welding, the arc length can be adjusted, and the welding seam quality is effectively improved. However, the circular seam direct current pulse TIG all-position welding is a multivariable coupling complex dynamic process, the circular seam welding arc length fluctuation is large, the current arc voltage tracking method has the defects of low acquisition frequency, poor tracking quality and stability and the like, and the welding quality requirement of a large-scale tube plate circular seam robot is difficult to meet.

Patent CN 100548558C discloses a pipe welding torch control device, which controls the arc length by controlling the arc length and automatically controlling the swing arm and the height between the welding torch and the pipe to be welded, but the device has a limited range of pipe diameter to be welded. Patent CN 101811212B discloses an arc length controller for electrogas welding, which is composed of a voltage negative feedback current positive feedback circuit, an optical coupling isolation circuit, an AD conversion circuit and an FPGA control circuit. Patent CN 210731316U discloses a method for controlling the operation of a stepping motor by feeding back the arc voltage to a welding power supply, so as to stabilize the arc length at a certain height, and the tracking effect of the method is greatly influenced by the performance of the welding machine. Patent CN200420036039.0 adopts singlechip and PWM circuit to calculate the deviation of arc length according to the arc voltage change to realize the adjustment of arc length, the high frequency high pressure that nevertheless the arcing produced has great influence to the steady operation of singlechip.

In view of the above, the invention discloses a real-time arc voltage tracking method based on an arc voltage module, so as to solve the problem of poor welding arc length consistency caused by large girth welding voltage fluctuation, tube plate thermal deformation, tungsten electrode abrasion and the like.

Disclosure of Invention

In order to overcome the defects of the prior art, the invention provides a real-time arc length control method for tube plate TIG welding.

The technical scheme adopted by the invention is as follows:

a novel real-time arc voltage tracking method for arc welding is characterized in that:

a calibration step: welding the circular seam of the tube plate, and collecting the arc voltage of direct current pulse TIG welding by an arc voltage module according to a set frequency; extracting peak arc voltage, carrying out S-G filtering, taking the average value after filtering as the representative arc voltage of direct current pulse TIG welding, and obtaining the upper boundary arc voltage and the lower boundary arc voltage of arc voltage tracking by adopting a statistical analysis method based on coefficient of variation evaluation on the representative arc voltage;

a real-time welding step: the arc voltage module collects arc voltage of direct current pulse TIG welding according to set frequency to obtain real-time arc voltage, the real-time arc voltage is input to a boundary arc voltage tracking model, the model respectively compares the real-time arc voltage with the boundary arc voltage in the calibration step, and the fuzzy controller selects different fuzzy control methods according to the numerical values of the real-time arc voltage and the boundary arc voltage and adjusts the arc voltage in real time after determining the arc voltage tracking direction and the step length.

In the calibration step, the filtered mean value is the peak value arc voltage extracted and S-G filtered, the calculated filtered peak value arc voltage mean value is used as the characterization arc voltage of direct current pulse TIG welding. The method for acquiring the upper boundary arc voltage and the lower boundary arc voltage comprises the following steps: selecting m calibration circular seams on the tube plate, solving the characterization arc pressure of the m circular seams, and performing statistical analysis on the characterization arc pressure of the m circular seams by adopting a coefficient of variation evaluation method, wherein the method specifically comprises the following steps: calculating the mean value of m characterization arc voltages, respectively subtracting the maximum value and the minimum value of the m characterization arc voltages from the mean value, taking the maximum difference value as a threshold value, respectively subtracting the sum and the sum of the mean value and the threshold value, and determining the upper boundary arc voltage V of arc voltage tracking_{Upper boundary}And lower boundary arc voltage V_{Lower boundary}。

In the above real-time arc voltage tracking method for arc welding, the method for acquiring the real-time arc voltage comprises: receiving peak signals, delaying a set time T, collecting the average value of n peak arc voltages at intervals of the set time T, and averaging the n peak arc voltages to obtain the peak value stageReal-time arc voltage V_{Real time}The peak signal indicates that the welding current enters the pulse peak phase.

In the real-time arc voltage tracking method for arc welding, the real-time arc voltage tracking specifically comprises the following steps:

after receiving the base value signal, V_{Real time}And V_{Upper boundary}And V_{Lower boundary}And comparing, calculating arc voltage errors and the change rate thereof, inputting the calculation result to a fuzzy controller, wherein the fuzzy controller comprises an upper arc voltage fuzzy controller and a lower arc voltage fuzzy controller, the difference value between the real-time arc voltage and the upper boundary arc voltage and the lower boundary arc voltage is the arc voltage error, the arc voltage error is divided by the set time t to be the arc voltage error change rate, and the base value signal represents that the welding current enters a pulse base value stage.

In the real-time arc voltage tracking method for electric arc welding, the fuzzy controller performs fuzzy control by using V_{Real time}And V_{Upper boundary}And V_{Lower boundary}And comparing, and performing selection execution according to a comparison result:

if the real-time arc voltage V_{Real time}Greater than or equal to upper boundary arc voltage V_{Upper boundary}Then the arc voltage error and the change rate of the arc voltage error and the change rate are used as the input of the upper arc voltage fuzzy controller;

if the real-time arc voltage V_{Real time}Lower boundary arc voltage V or less_{Lower boundary}And then the arc voltage error and the change rate of the arc voltage error and the change rate are used as the input of the lower arc voltage fuzzy controller.

If the real-time arc voltage V_{Real time}Greater than the lower boundary arc voltage V_{Lower boundary}And is less than the upper boundary arc voltage V_{Upper boundary}The trend of the welding gun does not need to be controlled.

In the above real-time arc voltage tracking method for arc welding, the control method of the upper stage arc voltage fuzzy controller is as follows:

defining the arc voltage error e of the upper arc voltage₁And its rate of change ec₁Membership function library u, control rule library R and clarification algorithm fd through respective quantization factor k_e1And k_ec1After zooming, inputting the input to the upper arc voltage fuzzyController, i.e. two-dimensional input. The fuzzification module D/F is used for quantifying the arc voltage error e₁And ec₁Respectively converted into fuzzy quantity E and EC, and the reasoning module M outputs fuzzy quantity U, and the sharpening module F/D converts the fuzzy quantity U into sharpening step length L₁Finally, the proportional factor k of the upper arc voltage_u1Zooming, calculating the tracking step length s of the upper arc voltage₁。

Wherein the quantization factor k_e1Is calculated by the step of calculating the arc voltage error e₁Has an actual value range of [ x_min，x_max]Error of arc voltage e₁The value range after fuzzification is [ y_min，y_max]Quantization factor k_e1The calculation formula is as follows: k is a radical of_e1＝(y_max－y_min)/(x_max－x_min) Similarly, the quantization factor k can be obtained_ec1. Scale factor k_u1The calculation step is that the range of the feeding step length output by the upper-section arc voltage fuzzy controller is [ l_min，l_max]Feed step length s₁Has an actual value range of [ s ]_min，s_max]Then the scale factor k_u1The calculation formula is as follows: k is a radical of_u1＝(s_max－s_min)/(l_max－l_min)。

The step length L of the sharpening calculated by the upper fuzzy controller₁By a scale factor k_u1After scaling, the actual feeding step s is obtained₁When the actuating mechanism adjusts the feeding of the welding gun, the welding gun is close to the tube plate, the arc length j between the welding gun and the tube plate is reduced, and the welding gun coordinate is changed into z_i+1。

In the real-time arc voltage tracking method for arc welding, the control method of the lower-stage arc voltage fuzzy controller is as follows:

defining arc voltage error e of lower segment arc voltage₂And its rate of change ec₂Membership function library u, control rule library R and clarification algorithm fd through respective quantization factor k_e2And k_ec2After zooming, the data is input into a lower arc voltage fuzzy controller, namely two-dimensional input. The fuzzification module D/F is used for quantifying the arc voltage error e₂And ec₂Are respectively provided withConverting into fuzzy quantity E and EC, outputting fuzzy quantity U by inference module M, converting fuzzy quantity U into sharpening step length L by sharpening module F/D₂Finally, the scale factor k of the lower segment arc voltage_u2Zooming, calculating the tracking step length s of the lower segment arc voltage₂。

Wherein the quantization factor k_e2Is calculated by the step of calculating the arc voltage error e₂Has an actual value range of [ x_min，x_max]Error of arc voltage e₂The value range after fuzzification is [ y_min，y_max]Quantization factor k_e2The calculation formula is as follows: k is a radical of_e1＝(y_max－y_min)/(x_max－x_min) Similarly, the quantization factor k can be obtained_ec2。

Scale factor k_u2The calculation step is that the value range of the output feeding step length of the lower arc pressure fuzzy controller is [ l ]_min，l_max]Feed step length s₂Has an actual value range of [ s ]_min，s_max]Then the scale factor k_u1The calculation formula is as follows: k is a radical of_u2＝(s_max－s_min)/(l_max－l_min)。

The definition step length L calculated according to the lower fuzzy controller₂By a scale factor k_u2After scaling, the actual feeding step s is obtained₂At the moment, the actuating mechanism adjusts the feeding of the welding gun, the welding gun is far away from the tube plate, the arc length j between the welding gun and the tube plate is increased, and the welding gun coordinate is changed into z_i+1。

Compared with the prior art, the invention has the beneficial effects that: the method can effectively adjust the welding arc length in real time, compensate the welding distance change caused by the welding deformation of the tube plate and avoid the generation of welding defects.

Drawings

FIG. 1 is a diagram of arc voltage tracking hardware components.

FIG. 2 shows peak arc voltage extraction and S-G filtering.

FIG. 3 is a boundary arc voltage tracking strategy.

Fig. 4 shows the basic structure of the Mamdani type fuzzy controller.

Fig. 5 is a two-dimensional three-segment fuzzy control structure.

FIG. 6 is an overall flow chart of arc voltage tracking.

Detailed Description

The invention relates to high-frequency acquisition of arc voltage signals, S-G filtering and characterization of arc voltage signals, a boundary arc voltage tracking strategy and a tracking step fuzzy control method. In the characterization stage, the arc voltage module collects peak arc voltage of direct current pulse TIG welding at a certain frequency, adopts an S-G smooth filtering method to set a filtering order and a sliding window width, processes collected peak voltage data and removes burrs of an arc voltage signal. And taking the filtered peak value arc voltage mean value as the characterization arc voltage of the direct current pulse TIG welding, performing statistical analysis on the characterization arc voltages of the plurality of circular seams by adopting a variation coefficient evaluation method, and determining the upper boundary arc voltage and the lower boundary arc voltage of the arc voltage tracking. In the tracking stage, a tracking strategy based on the boundary arc voltage is designed: the real-time arc voltage is calculated in the pulse peak value stage, the tungsten electrode position is adjusted in the pulse base value stage, and 1 pulse square wave is used for completing 1 real-time arc voltage calculation and 1 arc voltage tracking. A Mamdani type two-dimensional three-section fuzzy controller is designed, arc voltage quantization factors and scale factors of all sections are calculated, arc voltage errors and change rates of the arc voltage errors are used as input of fuzzy control, tracking step length is used as output of the fuzzy control, and real-time arc voltage tracking is achieved through feeding movement of a welding gun.

The invention is further described below with reference to the accompanying drawings.

Fig. 1 is a diagram of arc voltage tracking hardware, which mainly comprises an arc voltage module, an industrial personal computer, a motion control unit and the like. When the tungsten electrode is welded to the circular seam of the tube plate, the arc voltage module collects arc voltage signals at a set frequency, meanwhile, the arc voltage module calculates real-time arc voltage and compares the real-time arc voltage with the boundary arc voltage set by the industrial personal computer, the arc voltage error and the change rate of the arc voltage error are calculated through the arc voltage module and are input into the built-in fuzzy controller, the tracking step length and the tracking direction of the welding gun are determined, and the arc voltage is adjusted. Meanwhile, signal communication is also carried out between the motion control unit and the industrial personal computer, the industrial personal computer generates a G code to the motion control unit, the circumferential weld positioning of the welding gun is realized, and a register reading and writing function is also arranged between the motion control unit and the industrial personal computer.

FIG. 2 shows peak extraction and S-G filtering, and the arc voltage module collects arc voltage signals of direct current pulse TIG welding at 1000Hz, the pulse period is 400ms, and the duration of the peak value and the base value are 200ms respectively. Taking the arc voltage signal of 1 pulse as an example, extracting a peak signal to obtain 200 sampling points, processing the peak signal by adopting an S-G smooth filtering method, wherein the filtering order is 6 orders, the width of a sliding window is 399, extracting the peak signal for all the pulses on the basis of the S-G smooth filtering method, carrying out S-G smooth filtering, and further taking the average value of the filtered arc voltages as the representative arc voltage of the direct current pulse TIG welding. Selecting m calibration circular seams on a tube plate, solving the characterization arc pressure of the m circular seams, carrying out statistical analysis on the characterization arc pressure of the m circular seams by adopting a coefficient of variation evaluation method, solving the mean value of the m characterization arc pressures, respectively subtracting the maximum value and the minimum value of the m characterization arc pressures from the mean value, taking the maximum difference value as a threshold value, respectively subtracting the sum and the difference of the mean value from the threshold value, and determining the upper boundary arc pressure V of arc pressure tracking_{Upper boundary}And lower boundary arc voltage V_{Lower boundary}。

FIG. 3 shows a boundary arc voltage tracking strategy, in which an arc voltage module receives a peak signal and delays for 50ms to avoid arc voltage "glitches", then 1 peak arc voltage is collected every 1ms, n peak arc voltages are collected in total, the value of n is related to the peak time, and the real-time arc voltage V at the peak stage can be obtained by averaging the n peak arc voltages_{Real time}. After the arc voltage module receives the base value signal, the V is converted into the voltage_{Real time}V obtained as described above_{Upper boundary}And V_{Lower boundary}And comparing, calculating the arc voltage error and the change rate thereof (the difference value between the real-time arc voltage and the upper boundary arc voltage and the lower boundary arc voltage is the arc voltage error, dividing the arc voltage error by the set time t to be the change rate of the arc voltage error), and inputting the calculation result to a subsequent controller.

FIG. 4 is a basic structure of a Mamdani fuzzy controller, u represents a membership function library, R represents a control rule library, fd represents a sharpening algorithm, e and ec represent arc voltage errors and change rates thereof, and k_eAnd k_ecFor respective quantization factors, blurringThe quantization module D/F converts the quantized arc voltage errors E and EC into fuzzy quantities E and EC respectively, the reasoning module M outputs a fuzzy quantity U, the clarification module F/D converts the fuzzy quantity U into a clear quantity L, and the clear quantity L is subjected to a scale factor k_uAnd outputting a tracking step length s under the action of the control unit. The upper and lower sections include:

the control method of the upper-section arc voltage fuzzy controller comprises the following steps:

defining the arc voltage error e of the upper arc voltage₁And its rate of change ec₁Membership function library u, control rule library R and clarification algorithm fd through respective quantization factor k_e1And k_ec1After zooming, the input is input into an upper arc voltage fuzzy controller, namely a two-dimensional input. The fuzzification module D/F is used for quantifying the arc voltage error e₁And ec₁Respectively converted into fuzzy quantity E and EC, and the reasoning module M outputs fuzzy quantity U, and the sharpening module F/D converts the fuzzy quantity U into sharpening step length L₁Finally, the proportional factor k of the upper arc voltage_u1Zooming, calculating the tracking step length s of the upper arc voltage₁。

Wherein the quantization factor k_e1Is calculated by the step of calculating the arc voltage error e₁Has an actual value range of [ x_min，x_max]Error of arc voltage e₁The value range after fuzzification is [ y_min，y_max]Quantization factor k_e1The calculation formula is as follows: k is a radical of_e1＝(y_max－y_min)/(x_max－x_min) Similarly, the quantization factor k can be obtained_ec1. Scale factor k_u1The calculation step is that the range of the feeding step length output by the upper-section arc voltage fuzzy controller is [ l_min，l_max]Feed step length s₁Has an actual value range of [ s ]_min，s_max]Then the scale factor k_u1The calculation formula is as follows: k is a radical of_u1＝(s_max－s_min)/(l_max－l_min)。

The step length L of the sharpening calculated by the upper fuzzy controller₁By a scale factor k_u1After scaling, the actual feeding step s is obtained₁At the moment, the actuating mechanism adjusts the feeding of the welding gun, the welding gun is close to the tube plate,the arc length j between the torch and the tube sheet is reduced and the torch coordinate becomes z_i+1。

The control method of the lower arc voltage fuzzy controller comprises the following steps:

defining arc voltage error e of lower segment arc voltage₂And its rate of change ec₂Membership function library u, control rule library R and clarification algorithm fd through respective quantization factor k_e2And k_ec2After zooming, the data is input into a lower arc voltage fuzzy controller, namely two-dimensional input. The fuzzification module D/F is used for quantifying the arc voltage error e₂And ec₂Respectively converted into fuzzy quantity E and EC, and the reasoning module M outputs fuzzy quantity U, and the sharpening module F/D converts the fuzzy quantity U into sharpening step length L₂Finally, the scale factor k of the lower segment arc voltage_u2Zooming, calculating the tracking step length s of the lower segment arc voltage₂。

Wherein the quantization factor k_e2Is calculated by the step of calculating the arc voltage error e₂Has an actual value range of [ x_min，x_max]Error of arc voltage e₂The value range after fuzzification is [ y_min，y_max]Quantization factor k_e2The calculation formula is as follows: k is a radical of_e1＝(y_max－y_min)/(x_max－x_min) Similarly, the quantization factor k can be obtained_ec2。

Scale factor k_u2The calculation step is that the value range of the output feeding step length of the lower arc pressure fuzzy controller is [ l ]_min，l_max]Feed step length s₂Has an actual value range of [ s ]_min，s_max]Then the scale factor k_u1The calculation formula is as follows: k is a radical of_u2＝(s_max－s_min)/(l_max－l_min)。

The definition step length L calculated according to the lower fuzzy controller₂By a scale factor k_u2After scaling, the actual feeding step s is obtained₂At the moment, the actuating mechanism adjusts the feeding of the welding gun, the welding gun is far away from the tube plate, the arc length j between the welding gun and the tube plate is increased, and the welding gun coordinate is changed into z_i+1。

FIG. 5 is a two-dimensional three-segment moldAnd the fuzzy control structure divides the whole arc voltage interval into three sections, namely an upper section arc voltage (more than or equal to the upper boundary arc voltage), a middle section arc voltage (more than the lower boundary arc voltage and less than the upper boundary arc voltage) and a lower section arc voltage (less than or equal to the lower boundary arc voltage) according to the boundary arc voltage tracking strategy. Based on the above, a fuzzy controller can be respectively added for the upper arc voltage and the lower arc voltage, and the middle arc voltage does not need to be controlled according to a tracking strategy. Each fuzzy controller adopts a Mamdani type fuzzy controller structure, and the arc voltage error e of the upper arc voltage₁And its rate of change ec₁By respective quantization factor k_e1And k_ec1After zooming, the data is input into an upper arc voltage fuzzy controller, namely two-dimensional input, and the upper arc voltage fuzzy controller outputs a clear quantity step length L through a fuzzy control rule table₁Finally, the proportional factor k of the upper arc voltage_u1Calculating the tracking step length s of the upper arc voltage by zooming_{On the upper part}. Similarly, the fuzzy control of the lower segment arc pressure is similar to the above, and is not repeated. And the executing mechanism realizes real-time regulation of the arc voltage according to the tracking direction and the tracking step length of the fuzzy controller.

FIG. 6 is a flow chart of the real-time arc voltage tracking method according to the present invention, which includes selecting a calibration circular seam, collecting an arc voltage during a welding process, extracting a peak arc voltage of the arc voltage, obtaining a smooth arc voltage by an S-G filtering method, calculating a mean value of the calibrated circular seam smooth arc voltage, using the mean value as a DC pulse characterization arc voltage, repeating the steps to obtain m characterization arc voltages of the calibration circular seam, and performing statistical analysis on the m characterization arc voltages of the circular seam by a variation coefficient evaluation method to obtain an upper boundary arc voltage and a lower boundary arc voltage of the calibration circular seam. And (3) welding a target circular seam, extracting peak arc voltage at intervals of 1ms after receiving a peak signal, accumulating n peak arc voltages, calculating an average value, taking the average value as the real-time arc voltage of the pulse, comparing the real-time arc voltage with the boundary arc voltage according to a boundary arc voltage tracking strategy after receiving a base value signal, determining a tracking direction and a tracking step length according to upper-section and lower-section arc voltage fuzzy control respectively, and inputting the tracking direction and the tracking step length into a motion control unit to realize the adjustment of the arc voltage.

The specific embodiments described herein are merely illustrative of the spirit of the invention. Various modifications or additions may be made to the described embodiments or alternatives may be employed by those skilled in the art without departing from the spirit or ambit of the invention as defined in the appended claims.