Electrode dislocation detection method and system based on dual-magnetic-pole Hall effect sensor

Abstract

The invention discloses an electrode dislocation detection method and system based on a dual-magnetic-pole Hall effect sensor, and the method comprises the steps: carrying out the offset voltage and noise elimination of a Hall effect sensor through a rotating current offset elimination method and a chopping stabilization technology elimination method, and obtaining the dual-magnetic-pole Hall effect sensor; hall voltage acquisition processing is performed on the electrode based on a dual-magnetic-pole Hall effect sensor, and the contact area variation of the tip of the electrode is obtained; and performing feature extraction processing on the contact area variable quantity of the electrode tip through a neural network model, determining the dislocation quantity of the electrode, and comparing the dislocation quantity with a standard threshold dislocation quantity to obtain an electrode dislocation detection result. According to the invention, the accuracy of magnetic field measurement can be improved by eliminating offset voltage and noise, so that the detection accuracy of electrode dislocation is improved. The electrode dislocation detection method and system based on the double-magnetic-pole Hall effect sensor can be widely applied to the technical field of electrode dislocation detection.

Description

Technical Field

[0001] The present invention relates to the technical field of electrode misalignment detection, and particularly to an electrode misalignment detection method and system based on a dual-pole Hall effect sensor. Background Art

[0002] Resistance spot welding is a metal connection technology widely used in the automotive manufacturing industry, and is favored for its high joint quality, high production efficiency, few auxiliary processes, and easy automation. During the resistance spot welding process, the alignment state of the electrode tips has a significant impact on the mechanical properties of the welded joint. If the two electrode tips are not on the same straight line, the area through which the current passes during welding becomes smaller, and the size of the formed fusion core will also decrease, resulting in a reduction in the mechanical properties of the welded joint. When there is a certain angular offset between the axes of the two electrode tips, the contact between the electrode tip and the material to be welded has a sequence. The part that first comes into contact with the electrode tip melts and causes a deeper indentation depth in this part. When the welding current is too large, spatter is more likely to occur in this part, resulting in less metal inside the final fusion core and forming defects such as porosity shrinkage cavities. Excessive indentation depth of the electrode tip of the welded joint will also cause stress concentration in this part, reducing the strength and toughness of the welded joint. When the axes of the electrode tips are on the same straight line, i.e., in alignment, the plastic ring seals well during the welding process, and the molten pool metal is not easily ejected, avoiding the generation of welding spatter. The obtained welded joint has a good appearance and a perfect symmetric cross-section of the fusion core, and the mechanical properties of the welded joint are also better, with improved strength and toughness. Therefore, the alignment accuracy of the resistance spot welding electrode tips directly determines the welding quality of the welded joint.

[0003] In an actual production environment, due to mechanical fatigue or improper adjustment of the welding gun, the electrode tips may experience lateral offset, which will lead to instability during the welding process and a decrease in welding quality. First, most traditional detection methods are destructive experiments after welding and cannot real-time monitor the lateral offset of the electrode tips during the welding process. This means that before welding is completed, the electrode offset problem cannot be detected and adjusted in a timely manner, resulting in defects that may occur during the welding process not being corrected in a timely manner. Second, due to the lack of real-time monitoring means, even if the electrode offset problem is detected, it cannot be adjusted in a timely manner during the welding process. This may lead to the generation of welding defects, affecting the mechanical properties of the welded joint and increasing the workload and cost of rework and repair. Traditional detection methods may not be able to accurately detect small electrode offsets, resulting in some potential quality problems being overlooked. These problems not only affect the welding quality but may also have an adverse impact on subsequent production processes and product quality. Summary of the Invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide an electrode misalignment detection method and system based on a dual-pole Hall effect sensor, which can improve the accuracy of magnetic field measurement by eliminating offset voltage and noise, thereby improving the detection accuracy of electrode misalignment.

[0005] The first technical solution adopted by the present invention is: an electrode misalignment detection method based on a bipolar Hall effect sensor, comprising the following steps:

[0006] Eliminate the offset voltage and noise of the Hall effect sensor through the rotating current offset cancellation method and the chopper stabilization technique cancellation method to obtain a bipolar Hall effect sensor;

[0007] Collect and process the Hall voltage of the electrode based on the bipolar Hall effect sensor to obtain the change amount of the contact area at the tip of the electrode;

[0008] Extract the features of the change amount of the contact area at the tip of the electrode through a neural network model, determine the misalignment amount of the electrode, and compare it with the standard threshold misalignment amount to obtain the electrode misalignment detection result.

[0009] Further, the bipolar Hall effect sensor includes a rotating current source, a Hall element, a modulation circuit, a chopper amplifier circuit, a demodulation circuit, a comparison circuit, and an output control circuit, and the rotating current source, the Hall element, the modulation circuit, the chopper amplifier circuit, the demodulation circuit, the comparison circuit, and the output control circuit are connected in sequence.

[0010] Further, the step of eliminating the offset voltage and noise of the Hall effect sensor through the rotating current offset cancellation method and the chopper stabilization technique cancellation method to obtain a bipolar Hall effect sensor specifically includes:

[0011] Based on the rotating current source, eliminate the offset voltage of the Hall effect sensor through the rotating current offset cancellation method to obtain a Hall effect sensor after eliminating the offset voltage;

[0012] Based on the chopper amplifier circuit, eliminate the offset voltage noise of the Hall effect sensor through the chopper stabilization technique cancellation method to obtain a Hall effect sensor after eliminating the offset voltage noise;

[0013] Combine the Hall effect sensor after eliminating the offset voltage and the Hall effect sensor after eliminating the offset voltage noise to obtain a bipolar Hall effect sensor.

[0014] Further, the step of based on the rotating current source, eliminating the offset voltage of the Hall effect sensor through the rotating current offset cancellation method to obtain a Hall effect sensor after eliminating the offset voltage specifically includes:

[0015] Generate a vertical-phase Hall element bias voltage and a horizontal-phase Hall element bias voltage through the rotating current source;

[0016] Based on the vertical-phase Hall element bias voltage, with the supply current of the Hall element flowing from top to bottom, obtain the vertical-phase output voltage of the Hall element;

[0017] Based on the horizontal-phase Hall element bias voltage, with the supply current of the Hall element flowing from left to right, obtain the horizontal-phase output voltage of the Hall element;

[0018] Superimpose and average the vertical-phase output voltage and the horizontal-phase output voltage of the Hall element to obtain a Hall effect sensor with the offset voltage eliminated.

[0019] Furthermore, the step of performing offset voltage noise elimination processing on the Hall effect sensor by the chopper stabilization technology elimination method based on the chopper amplifier circuit to obtain a Hall effect sensor with offset voltage noise eliminated specifically includes:

[0020] Power on the Hall effect sensor to obtain the Hall voltage to be processed;

[0021] Based on the chopper amplifier circuit, generate a high-frequency square-wave modulation signal to modulate the Hall voltage to be processed to the high-frequency band to obtain a high-frequency modulation signal;

[0022] Superimpose the high-frequency modulation signal with the offset voltage and noise of the Hall effect sensor to obtain a mixed signal;

[0023] After amplifying the mixed signal, perform demodulation processing through the high-frequency square-wave modulation signal to obtain a baseband signal;

[0024] Filter the high-frequency noise and offset voltage from the baseband signal to obtain a Hall effect sensor with offset voltage noise eliminated.

[0025] Furthermore, the step of collecting the Hall voltage of the electrode by the bipolar Hall effect sensor to obtain the change amount of the contact area at the electrode tip specifically includes:

[0026] Power on the electrode to obtain the current density of the electrode;

[0027] According to Ampere's circuital law, determine the relationship between the magnetic field strength of the electrode and the current density of the electrode;

[0028] Collect the Hall voltage of the electrode by the bipolar Hall effect sensor to obtain the Hall voltage of the electrode;

[0029] Calculate the electrode contact area by combining the Hall voltage of the electrode and the relationship between the magnetic field strength of the electrode and the current density of the electrode to obtain the change amount of the contact area at the electrode tip.

[0030] Furthermore, the expression for calculating the electrode contact area is specifically as follows:

[0031]

[0032] In the above formula, V H represents the Hall voltage of the electrode, K represents the sensitivity coefficient, k represents the proportionality coefficient, and C a represents the contact area of the electrode tip, and I represents the magnitude of the current passing through the electrode.

[0033] Furthermore, the step of performing feature extraction processing on the change amount of the contact area of the electrode tip through the neural network model, determining the misalignment amount of the electrode, and comparing it with the standard threshold misalignment amount to obtain the electrode misalignment detection result specifically includes:

[0034] Performing feature extraction on the change amount of the contact area of the electrode tip through the neural network model to obtain the mean and variance of the magnetic field intensity;

[0035] Determining the misalignment amount of the electrode according to the mean and variance of the magnetic field intensity;

[0036] Collecting the Hall voltage of the standard electrode based on the bipolar Hall effect sensor to obtain the misalignment amount of the standard electrode as the standard threshold misalignment amount;

[0037] If the misalignment amount of the electrode is greater than the standard threshold misalignment amount, there is an axial misalignment problem with the electrode currently, and an audible and visual alarm is triggered;

[0038] If the misalignment amount of the electrode meets the range of the standard threshold misalignment amount, there is no axial misalignment problem with the electrode currently, and the electrode misalignment detection result is output.

[0039] The second technical solution adopted by the present invention is: an electrode misalignment detection system based on a bipolar Hall effect sensor, including:

[0040] The first module is used to eliminate the offset voltage and noise of the Hall effect sensor through the rotation current offset elimination method and the chopper stabilization technique to obtain a bipolar Hall effect sensor;

[0041] The second module is used to collect the Hall voltage of the electrode based on the bipolar Hall effect sensor to obtain the change amount of the contact area of the electrode tip;

[0042] The third module is used to perform feature extraction processing on the change amount of the contact area of the electrode tip through the neural network model, determine the misalignment amount of the electrode, and compare it with the standard threshold misalignment amount to obtain the electrode misalignment detection result.

[0043] The beneficial effects of the method and system of the present invention are as follows: The present invention eliminates the offset voltage and noise of the Hall effect sensor through the rotation current offset elimination method and the chopper stabilization technology elimination method, obtaining a bipolar Hall effect sensor. By combining the rotation current offset elimination technology and the chopper stabilization technology, the offset voltage and noise are effectively eliminated, ensuring a more accurate output signal when the magnetic field changes. The rotation current offset elimination technology and the chopper stabilization technology are used to effectively reduce the offset and noise of the Hall element and the amplifier circuit, thereby improving the accuracy of magnetic field measurement. Further, based on the bipolar Hall effect sensor, the Hall voltage of the electrode is collected and processed to obtain the change amount of the contact area at the tip of the electrode. Finally, the neural network model is used to perform feature extraction processing on the change amount of the contact area at the tip of the electrode, determine the misalignment amount of the electrode, and compare it with the standard threshold misalignment amount, so that the electrode misalignment during the welding process can be monitored in real time, providing timely and accurate data support for the evaluation of welding quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 FIG. is a flowchart of the steps of the electrode misalignment detection method based on the bipolar Hall effect sensor of the present invention;

[0045] Figure 2 FIG. is a structural block diagram of the electrode misalignment detection system based on the bipolar Hall effect sensor of the present invention;

[0046] Figure 3 FIG. is a schematic diagram of the electrode alignment state provided by a specific embodiment of the present invention;

[0047] Figure 4 FIG. is a schematic diagram of the reduction of the contact area caused by axial misalignment provided by a specific embodiment of the present invention;

[0048] Figure 5 FIG. is a schematic diagram of the magnetic field generation principle provided by a specific embodiment of the present invention;

[0049] Figure 6 FIG. is a schematic diagram of the Wien bridge model provided by a specific embodiment of the present invention;

[0050] Figure 7 FIG. is a schematic diagram of the principle of the rotation current offset elimination technology provided by a specific embodiment of the present invention;

[0051] Figure 8 FIG. is a schematic diagram of the working process of the chopper stabilization technology provided by a specific embodiment of the present invention;

[0052] Figure 9 FIG. is a schematic diagram of the principle of the chopper stabilization technology provided by a specific embodiment of the present invention;

[0053] Figure 10 FIG. is a schematic diagram of the electrode offset detection process provided by a specific embodiment of the present invention;

[0054] Figure 11 It is a schematic diagram of the circuit framework of the dual-pole Hall effect sensor provided by a specific embodiment of the present invention;

[0055] Figure 12 It is a schematic diagram of the encapsulation of the dual-pole Hall effect sensor provided by a specific embodiment of the present invention;

[0056] Figure 13 It is a schematic diagram of the electrode pair alignment and misalignment monitoring system device provided by a specific embodiment of the present invention;

[0057] Figure 14 It is a schematic diagram of the process of electrode pair alignment and misalignment monitoring provided by a specific embodiment of the present invention;

[0058] Figure 15 It is a schematic diagram of the data collected during the sending inspection provided by a specific embodiment of the present invention. Specific Embodiments

[0059] The present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. For the step numbers in the following embodiments, they are only set for the convenience of explanation and illustration, and no limitation is imposed on the order between the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0060] First of all, it should be noted that the Hall effect sensor is a magnetic field sensor based on the Hall effect principle, which can convert the magnetic field signal into an electrical signal. With the progress of semiconductor technology, its performance has been continuously improved and its application scope has become increasingly wide. The dual-pole Hall effect sensor can detect the magnetic fields of both S and N poles, and has the characteristics of high sensitivity and high precision, and is suitable for applications where high requirements for magnetic field detection are required. In Hall sensors, the offset voltage and noise will affect the measurement accuracy. The development of rotating current offset cancellation technology and chopper stabilization technology has effectively reduced the offset voltage and noise of the sensor and improved the measurement accuracy. At the same time, with the continuous improvement of the power consumption requirements of electronic devices, low-power design technology has also been widely applied in the sensor field. By optimizing the circuit design and working mode, the power consumption can be significantly reduced on the premise of ensuring the performance of the sensor, and it can be used to detect the alignment or misalignment state of the electrode pairs in resistance welding.

[0061] Based on this, the embodiments of the present invention aim to provide a method and system for detecting electrode alignment and misalignment during resistance spot welding. First, a bipolar Hall effect sensor manufactured using COMS technology is designed, and a rotating current offset cancellation technique and a chopper stabilization technique are introduced to eliminate offset voltage noise, so as to obtain a more accurate output signal when the magnetic field changes. Four Hall effect sensors are configured in an anti-parallel layout to achieve synchronous acquisition of magnetic field data on four Cartesian coordinate axes to comprehensively capture magnetic field changes. Secondly, a data acquisition and communication protocol is designed. The data is transmitted from the sensor to the PLC through the LoRaWAN communication protocol and then stored in the database, realizing efficient processing and analysis of the data. Finally, a decision tree model is introduced to achieve real-time and non-destructive detection of electrode misalignment, timely evaluate the welding quality, and set a threshold through the result of the decision tree model. When the threshold is exceeded, an alarm is issued to timely adjust the welding parameters, improve the welding efficiency, reduce the production cost, promote the intelligent development of welding technology, and improve the stability and welding quality of the welding process.

[0062] Referring to Figure 1 , the present invention provides a method for detecting electrode misalignment based on a bipolar Hall effect sensor, and the method includes the following steps:

[0063] S100. Eliminate the offset voltage and noise of the Hall effect sensor through the rotating current offset cancellation method and the chopper stabilization technique to obtain a bipolar Hall effect sensor;

[0064] First of all, it should be noted that the embodiments of the present invention are based on the detection of magnetic field changes caused by electrode offset, and a bipolar Hall effect sensor manufactured using COMS technology is designed. The internal of this sensor eliminates offset voltage noise by introducing a rotating current offset cancellation technique and a chopper stabilization technique to obtain a more accurate output signal when the magnetic field changes.

[0065] Furthermore, it should also be noted that this design uses a bipolar Hall effect sensor manufactured based on CMOS technology and proposes a method for detecting the electrode alignment state by measuring the magnetic field on the electrode contact plane. The specific measurement process is as follows: Magnetic field measurements are respectively carried out on two perpendicular Cartesian coordinate axes (positive axis and negative axis), and the alignment of the electrodes is accurately judged by comparing the magnetic field differences generated by misaligned electrodes and perfectly aligned electrodes.

[0066] The embodiments of the present invention use a bipolar Hall effect sensor manufactured using COMS technology, as Figure 11 shown. One part is the main path of the signal, which includes a rotating current source, a Hall element, a modulation circuit, a chopper amplifier circuit, a demodulation circuit, a comparison circuit, and an output control circuit.

[0067] First, the rotation current source rotates the current. By changing the current direction of the Hall element, the offset voltage has opposite polarities at different phases, thus canceling out the offset voltage, and the Hall induction voltage remains unchanged. Therefore, as long as the rotation frequency of the current is much higher than the magnetic field change frequency, the output voltages of the two phases are superimposed and then averaged to eliminate the offset voltage and obtain a more accurate Hall induction signal. After processing the Hall induction signal through modulation, amplification, demodulation, comparison, etc., the best threshold is obtained through comparison and judgment, and finally the digital high and low levels are output after judging the threshold.

[0068] Another part is the auxiliary module of the circuit. The main component modules include the offset voltage compensation circuit, voltage stabilization circuit, Hall element supply current bias, sleep clock control circuit, etc., to assist in processing the main path signal. The functions of each circuit are explained as follows:

[0069] 1) Offset voltage compensation circuit: used to compensate the offset voltage of the Hall element during the rotation of the current to ensure the accuracy of the output signal

[0070] 2) Voltage stabilization circuit: The voltage stabilization circuit is used to convert the input power supply voltage with a wide input range into an accurate and constant standard voltage, provide a stable power supply for other modules inside the chip, suppress power supply noise, and reduce the impact of power supply fluctuations on the operation of the internal circuit of the chip.

[0071] 3) Hall element supply current bias: This circuit block uses a current with a certain temperature coefficient as the bias current source for the constant current supply of the Hall element to achieve voltage induction of the Hall element.

[0072] 4) Sleep clock control circuit: The integrated oscillator module charges and discharges the capacitor using a constant current to generate a clock signal and provides it to the remaining modules, controlling the sleep and working state switching of the overall circuit to achieve a low-power working mode.

[0073] In the digital switch Hall sensor chip according to the embodiment of the present invention, since there are only three pins, namely the power supply, reference ground, and output, its package diagram is as Figure 12 shown.

[0074] S110. Based on the rotation current source, the Hall effect sensor is processed to eliminate the offset voltage through the rotation current offset elimination method to obtain the Hall effect sensor after eliminating the offset voltage;

[0075] Specifically, a Hall element bias voltage of the vertical phase and a Hall element bias voltage of the horizontal phase are generated by a rotating current source; based on the Hall element bias voltage of the vertical phase, the supply current of the Hall element flows from top to bottom, and the vertical phase output voltage of the Hall element is obtained; based on the Hall element bias voltage of the horizontal phase, the supply current of the Hall element flows from left to right, and the horizontal phase output voltage of the Hall element is obtained; the vertical phase output voltage of the Hall element and the horizontal phase output voltage of the Hall element are superimposed and averaged to obtain a Hall effect sensor after eliminating the offset voltage.

[0076] First, the principle of the rotating current offset cancellation technique is described. A Hall element can usually be equivalently modeled as a Wien bridge. The distributed resistances of the four ports of the Hall element can be equivalently modeled as resistance elements R1 to R4 connected between adjacent ports, as Figure 6 shown. In the ideal equivalent Wien bridge model of the Hall element, terminals A, B, C, and D correspond to the four ports of the Hall element, and the distributed resistances R1 = R2 = R3 = R4. Assume that a bias voltage of magnitude V b is applied between ports A and C. In the ideal case, the potentials at terminals B and D are equal, both being V b , and there is no voltage difference. At this time, the offset voltage is zero, that is, V os = 0. However, in actual Hall elements produced, due to external factors such as the piezoresistive effect and manufacturing process differences, the resistance values of R1 to R4 are often not equal. Assume that R2 = R3 = R4 = R, and the resistance value of R1 has a deviation of R1 = R + ΔR. In the state of zero magnetic field, there will be an offset voltage V os between terminals D and B, and its expression is:

[0077]

[0078] To avoid the influence of the offset voltage on the result, an embodiment of the present invention proposes a rotating current offset cancellation technique. Only one Hall element needs to be used to eliminate the influence brought by the offset voltage.

[0079] As Figure 7 shown, through a set of complementary clock signals CLK and CLKB, it is used to "rotate" to generate a Hall element bias voltage V Figure 7 of 90° vertical phase (as shown in (a) of Figure 7 ) and 0° horizontal phase (as shown in (b) of b ). In the case of the 90° vertical phase, the supply current of the Hall element flows from top to bottom, and the output voltage V O can be expressed as:

[0080]

[0081] Further, in the case of the 0° horizontal phase, the supply current of the Hall element flows from left to right, and the output voltage V o can be expressed as:

[0082]

[0083] It can be seen that the polarities of the offset voltages of the Hall element in the two-phase cases are opposite, and the Hall induction voltage remains unchanged. Therefore, as long as it is ensured that the frequency of the current rotation far exceeds the frequency of the magnetic field change, the offset voltage can be eliminated by superimposing the output voltages of the two phases and taking the average value. At this time, the output voltage is:

[0084] V O = V H | 0° = V H | 90°

[0085] Adopting the rotating current offset cancellation technology can reduce the equivalent offset magnetic field to within ±30 μT at most, and can better achieve the effect of offset cancellation.

[0086] S120. Based on the chopper amplifier circuit, the offset voltage noise of the Hall effect sensor is eliminated by the chopper stabilization technology, and the Hall effect sensor after eliminating the offset voltage noise is obtained;

[0087] Specifically, the Hall effect sensor is powered on to obtain the Hall voltage to be processed; based on the chopper amplifier circuit, a high-frequency square-wave modulation signal is generated to modulate the Hall voltage to be processed to the high-frequency band, and a high-frequency modulation signal is obtained; the high-frequency modulation signal is superimposed with the offset voltage and noise of the Hall effect sensor to obtain a mixed signal; after the mixed signal is amplified, it is demodulated by the high-frequency square-wave modulation signal to obtain a baseband signal; the high-frequency noise and offset voltage of the baseband signal are filtered to obtain the Hall effect sensor after eliminating the offset voltage noise.

[0088] In this embodiment, the process of the chopper stabilization technology is as shown in Figure 8 . First, the measured Hall voltage is used as the input signal V in , and a high-frequency square-wave signal with a frequency of f chop is added as the modulation signal m1(t). The V in is modulated to the high-frequency band through a multiplier to generate a modulation signal V2. The modulated signal V2 is superimposed with the inherent offset voltage V os and noise V noise of the circuit to form a mixed signal, and the signal is amplified through an amplifier to obtain V amp . Subsequently, a square-wave signal m2(t) with the same frequency and phase as m1(t) is used as the demodulation signal to demodulate the high-frequency signal back to the baseband (the baseband signal is Vdemod ) After demodulation, the high-frequency noise and offset voltage are filtered out by a low-pass filter to retain the baseband signal and output a pure and useful signal V out , which contains the original low-frequency information.

[0089] The modulation process is further described. The input voltage is processed by chopper stabilization technology to eliminate offset voltage noise. This technology is based on the modulation and demodulation principle. By separating the useful signal and noise in the frequency domain, the stable amplification of the signal is achieved. Chopper stabilization technology is often used in conjunction with the rotating current technology and is widely applied in Hall sensors to effectively eliminate offset voltage and 1 / f noise. Its principle is as Figure 9 shown.

[0090] V os and V noise represent the offset voltage and noise of the circuit respectively. The chopper clock signals m1(t) and m2(t) are both square waves with a frequency of f chop . At the same time, in order to ensure that the input signal does not overlap with signals such as offset voltage and noise, the frequency of the input signal should be limited to less than half of f chop .

[0091] Assume that the period of the chopper clock signal is T, and the input signal is an analog quantity that changes with time. The square wave modulation signal m(t) is expanded according to the Fourier series, and its expression is:

[0092]

[0093] where C k represents the Fourier coefficient of the k-th term and the even terms are 0, which is expressed as:

[0094]

[0095] Therefore, the input signal V in is converted into odd harmonics after being modulated by the modulation signal m1(t), amplified by the operational amplifier, and then demodulated by the modulation signal m2(t). Its expression is:

[0096]

[0097] When k, l = 2n (n = 1, 2, 3...), V2 = 0.

[0098] After the input signal passes through the two-square-wave-signal modulation process, the signal spectrum will finally be located at 2nf chop . After the first modulation, the offset voltage V os and noise V noise are introduced, and only one modulation process is passed. Let the power spectral density of the offset and noise be S N(f), after being modulated by the m2(t) signal, we get:

[0099]

[0100] It can be seen that the modulated noise and offset only retain components at odd - multiple harmonic frequencies. The input signal after two - stage modulation processing is demodulated back to the baseband. The odd - harmonic frequencies of the chopping frequency become the frequency bands where the offset voltage and noise modulation are shifted. Finally, a low - pass filter is used to filter out the modulated noise and offset voltage.

[0101] S130. Combine the Hall - effect sensor with the offset voltage eliminated and the Hall - effect sensor with the offset - voltage noise eliminated to obtain a bipolar Hall - effect sensor.

[0102] S200. Based on the bipolar Hall - effect sensor, perform Hall - voltage acquisition processing on the electrode to obtain the change amount of the contact area at the electrode tip;

[0103] Specifically, energize the electrode to obtain the current density of the electrode; according to Ampere's circuital law, determine the relationship between the magnetic field strength of the electrode and the current density of the electrode; perform Hall - voltage acquisition processing on the electrode through the bipolar Hall - effect sensor to obtain the Hall voltage of the electrode; combine the Hall voltage of the electrode and the relationship between the magnetic field strength of the electrode and the current density of the electrode to calculate the contact area of the electrode, and obtain the change amount of the contact area at the electrode tip.

[0104] First of all, it should be noted that as Figure 3 shown, due to mechanical fatigue, the electrode loses its working axis and is displaced by δ relative to its reference axis. As shown in (b) of Figure 3 , electrode misalignment will cause adverse characteristics in the welding process and the quality of the welded joint. Misalignment or non - centering, whether axial or angular, may lead to irregular weld shapes and a reduction in weld size and projection due to the asymmetric distribution of forces and changes in the contact surface.

[0105] When the electrodes are misaligned, the contact area C at the electrode tip a changes according to the contact - surface equation formula, and its expression is:

[0106]

[0107] In the above formula, r represents the radius of the electrode tip, δ represents the misalignment between the electrodes, and C a represents the contact area at the electrode tip.

[0108] Therefore, the alignment of the welding electrodes has a significant impact on the quality of the spot - welded joint. The deformation of the contact surface is caused by misalignment, as shown in Figure 4For the cross-section of the tips of two electrodes when the electrodes are misaligned, it can be obtained from the figure and the formula that the reduction of the contact area is closely related to the electrode offset. When there is no electrode offset, that is, when δ = 0, the contact area is the largest (C a0 = πr 2 ), and as δ increases, C a decreases non-linearly.

[0109] Furthermore, the embodiment of the present invention proposes an initial method for detecting electrode offset by measuring the change in magnetic field. First, the theoretical influence between the electrode arrangement state and the magnetic field generated by it is studied. By simplifying the electrode into a wire with a constant current density, the magnetic field behavior generated by this wire follows Ampere's law. From Ampere's integral law, the relationship between the magnetic field (B) and the current density can be obtained. This law relates the magnetic field strength to its source (current density). Among them, the expression of Ampere's integral law is:

[0110]

[0111] In the above formula, B represents the magnetic field, J represents the current density, and E represents the electric field strength.

[0112] According to the expression of Ampere's integral law, it can be obtained that if current flows through the electrode, there is a current density, then a rotating magnetic field will appear around the electrode, and the rotor direction of the magnetic field is the same as the direction of the current density. This behavior is as Figure 5 shown.

[0113] According to the above introduction, when the welding current (I) is constant, when the electrode misalignment causes C a to decrease, the current density J increases significantly. Therefore, the current density (J) is inversely proportional to the contact area (C a ), and its relationship can be expressed as:

[0114]

[0115] According to Ampere's circuital law, the relationship between the magnetic field strength (B) and the current density (J) under the simplified model can be expressed as:

[0116]

[0117] In the above formula, k is a proportionality coefficient.

[0118] It is determined by the electrode geometry, material properties, and magnetic field distribution experiment calibration. This formula shows that the magnetic field strength B is linearly and positively correlated with the current density J (i.e., the reciprocal of the contact area C a ). During the detection process, k needs to be corrected repeatedly through data fitting to improve the accuracy of the model.

[0119] The output voltage of the Hall sensor (VH ) The relationship with the magnetic field strength (B) is:

[0120] V H = S·B·cosθ

[0121] In the above formula, S is the Hall coefficient, and θ is the angle between the magnetic field direction and the sensor plane.

[0122] After eliminating the angle influence through the calibration experiment, it is simplified to V H = K·B, where K is the comprehensive sensitivity coefficient. Combining with the magnetic field strength formula, finally we get:

[0123]

[0124] It can be concluded that the Hall voltage V H is inversely proportional to the contact area C a By measuring V in real time H the change of the contact area can be indirectly reflected, and thus the electrode misalignment δ can be deduced.

[0125] S300. Feature extraction processing is performed on the change amount of the contact area at the electrode tip through a neural network model to determine the electrode misalignment and compare it with the standard threshold misalignment to obtain the electrode misalignment detection result.

[0126] Specifically, feature extraction is performed on the change amount of the contact area at the electrode tip through a neural network model to obtain the mean and variance of the magnetic field strength; based on the mean and variance of the magnetic field strength, the electrode misalignment is determined; the Hall voltage of the standard electrode is collected through a bipolar Hall effect sensor to obtain the misalignment of the standard electrode as the standard threshold misalignment; if the misalignment of the electrode is greater than the standard threshold misalignment, there is an axial misalignment problem with the electrode currently, and an audible and visual alarm is triggered; if the misalignment of the electrode meets the range of the standard threshold misalignment, there is no axial misalignment problem with the electrode currently, and the electrode misalignment detection result is output.

[0127] First, as Figure 10 shown, the process of electrode offset detection is as follows: after the system starts, the bipolar Hall sensor is initialized first, and the rotation current offset cancellation technology and chopper stabilization technology are activated to eliminate noise, and the four-axis anti-parallel layout is calibrated; then a magnetic field is generated by energizing the electrode, and the Hall voltage V H is collected in real time, and the change of the contact area is calculated in combination with the formula ; after the data is received by the ESP32 microcontroller, it is transmitted to the PLC gateway through the LoRaWAN protocol and stored in the MySQL database; the decision tree model extracts features such as the mean and variance of the magnetic field strength to judge whether the misalignment δ exceeds the threshold δ max , if it exceeds the limit, an audible and visual alarm is triggered and the data is recorded. If the misalignment is within the threshold, it enters the low-power sleep mode to wait for the next round of detection.

[0128] Furthermore, as Figure 13 shown, integrating the Hall sensor with the electrode cap grinder can be fixed on the side of the electrode cap grinder. After a certain number of welding points are welded, the electrode cap is ground on the electrode cap grinder. After the grinding of the electrode cap is completed, the upper and lower electrodes are respectively moved to directly above and directly below the center of the PCB circumference, as Figure 14 shown. The electrode misalignment detection of the welding robot arm is the same. Control the upper and lower electrodes to make contact, design a preset current to trigger the generation of a magnetic field, and then use four Hall sensors to collect magnetic field data and convert this value into a voltage, and set a threshold by comparing it with a set of electrode centering. The data acquisition and signal processing module is used to calculate the average magnetic field of each sensor for each inspection and send the difference between the axes to the welding database for subsequent data processing and judgment.

[0129] As Figure 15 shown, for the communication protocol between the device and the database, after the microcontroller based on ESP32 receives the information, it obtains the magnetic field value, calculates the magnetic field difference of each axis, and then sends the value to the PLC. The device collects data and sends it to the gateway through the LoRaWAN communication protocol, and the gateway then sends the data to the PLC. Finally, the PLC serves as the gateway between the LoRa gateway and the database for storing data for analysis.

[0130] In summary, the specific implementation of the embodiments of the present invention is described as follows:

[0131] 1) Selection and installation of Hall effect sensors.

[0132] Select a bipolar Hall effect sensor manufactured based on CMOS technology, which integrates rotating current offset cancellation technology and chopper stabilization technology inside, can effectively eliminate offset voltage noise, and improve the accuracy of magnetic field measurement. At the same time, considering the temperature stability of the sensor, select a Hall effect sensor with low temperature drift characteristics to reduce the influence of temperature changes on the magnetic field measurement results. Install the Hall effect sensor near the electrode of the resistance spot welding equipment, ensure that the measurement plane of the sensor is parallel to the contact plane of the electrode, and the center of the sensor is located on the electrode axis to accurately measure the magnetic field generated by the electrode. The sensor is fixed on the bracket of the equipment by screws to ensure its stability during the welding process. The installation position should avoid being affected by factors such as welding current and electromagnetic interference. In order to achieve the acquisition of multi-directional magnetic field data, configure the sensors in an anti-parallel layout, which can synchronously collect magnetic field data on four Cartesian coordinate axes.

[0133] 2) Microcontroller and communication module.

[0134] The ESP32 microcontroller is adopted, which has powerful data processing capabilities and multiple communication interfaces. The ESP32 is connected to the Hall effect sensor through the SPI interface to achieve rapid acquisition of sensor data. An LoRaWAN communication module is installed on the ESP32 microcontroller to send the collected magnetic field data to the gateway. An industrial gateway that supports the LoRaWAN protocol is selected, which can receive the LoRaWAN data packets sent from the ESP32 microcontroller and convert them into a signal format suitable for PLC processing. The gateway is connected to the PLC through Ethernet. The PLC serves as a data transfer station, stores the magnetic field data in the database, and makes alarm judgments based on the set thresholds. Secondly, a MySQL database is built to store the magnetic field data received from the PLC and the alarm records. Corresponding data tables are created in the database, including the magnetic field data table, the alarm record table, etc., to effectively manage and query the data.

[0135] 3) Integration of decision tree algorithms.

[0136] Extract key features from the magnetic field data collected by the Hall effect sensor, including magnetic field strength, rate of change, and other relevant statistical parameters (such as mean, variance, peak value, etc.). Map these features to the geometric parameters of the radar chart, and define clear dimensions for each feature in the radar chart, so that these geometric parameters can intuitively represent the alignment state and misalignment degree of the electrodes.

[0137] Use the labeled training data set containing magnetic field features under different electrode alignment states and the corresponding alignment states to apply the decision tree algorithm to build a model. The decision tree recursively divides the feature space to form a series of decision rules for classifying the alignment state of the electrodes. Use methods such as cross-validation to strictly train and evaluate the constructed decision tree model to ensure the accuracy and robustness of the model. By adjusting the parameters of the decision tree (such as the depth of the tree, splitting criterion, etc.), optimize the model performance and improve its ability to distinguish normal alignment and misalignment states.

[0138] 4) Real-time monitoring and alarm.

[0139] In practical applications, the real-time collected magnetic field features are input into the trained decision tree model. The model quickly outputs a judgment on the electrode alignment state based on the input feature data. Combine the output of the decision tree model to set the alarm threshold. When the system detects electrode misalignment, it automatically issues an alarm and records the relevant data, providing a basis for subsequent analysis and optimization.

[0140] 5) Development of database management and query interface.

[0141] Develop a web-based database management and query interface. Build the front-end page using PHP and HTML technologies, and interact with the MySQL database through the MySQLi extension. The front-end page adopts a responsive design, which can adapt to different terminal devices and provide a good user experience. Use CSS style sheets to beautify the page, making its interface simple, intuitive, and easy to use.

[0142] In the management and query interface, provide functions such as magnetic field data query, alarm record query, and data statistical analysis. Users can quickly retrieve the required data by entering query conditions, such as time range, electrode number, magnetic field strength range, etc. The data statistical analysis function can statistically analyze historical data and display it in an intuitive chart form, such as bar charts, line charts, pie charts, etc., providing decision-making support for users. Set user permission management, and different users have different operation permissions to ensure the security and confidentiality of data.

[0143] Therefore, the embodiments of the present invention are different from the prior art in the following aspects:

[0144] 1) The bipolar Hall effect sensor manufactured based on CMOS technology can replace the traditional continuous working form with a sampling-sleep working form. When designing the sensor, the rotating current offset cancellation technology and chopper stabilization technology are used to eliminate the offset voltage and noise. The wide temperature range of this sensor is from -40°C to 125°C, and the sensor supports the detection of magnetic fields with dual magnetic polarity switch types.

[0145] 2) It can real-time monitor the lateral offset of the electrode pressure during the resistance spot welding process, and timely judge the electrode alignment and misalignment status, so as to be able to timely repair and adjust the electrode to ensure the stability of the welding process and the welding quality.

[0146] 3) Use four Hall effect sensors configured in an anti-parallel layout to achieve synchronous acquisition of magnetic field data on four Cartesian coordinate axes to comprehensively capture magnetic field changes.

[0147] 4) By measuring the magnetic field and correlating it with the electrode misalignment status, a new monitoring and correction method is provided, which does not rely on traditional destructive testing.

[0148] 5) Design a data acquisition and communication protocol. Through the LoRaWAN communication protocol, transmit data from the sensor to the PLC and then store it in the database, realizing the efficient processing and analysis of data.

[0149] 6) Seamlessly integrate components such as Hall effect sensors, microcontrollers, communication modules, PLCs, databases, and decision tree algorithms to form a complete monitoring system for resistance spot welding electrode alignment and misalignment, realizing real-time monitoring and accurate judgment of the electrode alignment status.

[0150] 7) Extract key features from the magnetic field data collected by the Hall effect sensor, including magnetic field intensity, rate of change, and other relevant statistical parameters, and map these features to the geometric parameters of the radar chart to visually represent the alignment state and misalignment degree of the electrodes.

[0151] 8) Utilize a labeled training data set containing magnetic field features under different electrode alignment states and the corresponding alignment states, and apply the decision tree algorithm to build a model. By recursively splitting the feature space, a series of decision rules are formed for classifying the alignment state of the electrodes.

[0152] 9) The magnetic field features collected in real time are input into the trained decision tree model. The model quickly outputs a judgment on the electrode alignment state based on the input feature data, and an alarm threshold is set in combination with the output of the decision tree model to achieve automatic alarm and data recording.

[0153] Therefore, the embodiments of the present invention have the following advantages compared with the prior art:

[0154] 1) A bipolar Hall effect sensor manufactured using CMOS technology, combined with rotating current offset cancellation technology and chopper stabilization technology, effectively eliminates offset voltage and noise, ensuring a more accurate output signal when the magnetic field changes. At the same time, the anti-parallel layout of the four Hall effect sensors can synchronously collect magnetic field data on the four Cartesian coordinate axes, comprehensively capturing magnetic field changes and further improving the detection accuracy.

[0155] 2) The designed data acquisition and communication protocol uses the LoRaWAN communication protocol to transmit sensor data to the PLC and then store it in the database, realizing efficient data processing and analysis. This enables real-time monitoring of electrode misalignment during the welding process, providing timely and accurate data support for the evaluation of welding quality.

[0156] 3) Introduce a decision tree model to achieve real-time and non-destructive detection of electrode misalignment. By setting a threshold, when the detection result exceeds the threshold, the system can promptly issue an alarm to remind the operator or the automation system to adjust the welding parameters, avoiding the generation of welding defects and improving welding efficiency and quality.

[0157] 4) By real-time monitoring and timely adjustment of welding parameters, reduce welding defects and rework caused by electrode offset, reduce material waste and repair costs during the production process, and thus effectively reduce the overall production cost.

[0158] 5) This embodiment combines Hall effect sensor technology, data communication technology, and intelligent monitoring technology to achieve intelligent monitoring and control of the welding process. This helps to promote the development of welding technology towards a more efficient and intelligent direction, improving the technical level and production efficiency of the entire welding industry.

[0159] 6) By real-time monitoring and timely adjustment of the electrode misalignment, ensure the accurate alignment of the electrodes during the welding process, improve the mechanical properties of the welded joint, enhance the stability and reliability of the welding process, and reduce the occurrence of welding defects and failures.

[0160] In summary, in the embodiments of the present invention, the rotation current offset cancellation technology and the chopper stabilization technology are first used to effectively reduce the offset and noise of the Hall element and the amplifier circuit, thereby improving the accuracy of magnetic field measurement. At the same time, aiming at the error of the traditional Hall element model, an optimized structural model is proposed, and the relevant parameters of the Hall element are accurately set through a hardware description language to ensure the accuracy of the simulation design and further improve the detection accuracy. Furthermore, by measuring the magnetic field of the welding torch and obtaining the magnetic field values on the Cartesian axes, the magnetic field distribution under the electrode alignment state can be more comprehensively and accurately reflected, providing rich data support for accurately detecting electrode alignment and misalignment. Combining a reliable alarm sending protocol and the alarm threshold determined by experiments, it is possible to timely and accurately judge whether the electrode is misaligned, avoid the generation of welding defects, and finally, through the decision tree model, complex pattern recognition and classification can be performed based on the data collected by the sensor, improving the detection accuracy of electrode misalignment and reducing the situations of misjudgment and missed judgment.

[0161] Furthermore, a working mode of switching between sleep and sampling is also adopted. On the premise of ensuring the performance of the sensor, the power consumption is greatly reduced, enabling the sensor to work stably for a long time, providing a basis for real-time monitoring. The system can collect the magnetic field data of the welding torch in real time and send the data to the data analysis software through the LoRa communication protocol to realize real-time monitoring and data analysis of the welding process, timely discover problems in the welding process and make adjustments. The decision tree model can quickly process the data collected in real time to realize real-time evaluation of electrode misalignment, providing strong support for real-time monitoring. And by establishing a reliable alarm sending protocol and determining a reasonable alarm threshold, when it is detected that the electrode misalignment exceeds the set threshold, the system can issue an alarm in time to remind the operator or automatically adjust the welding parameters, avoid the generation of welding defects, improve the welding quality and production efficiency. Based on the evaluation results of the decision tree model, it can quickly judge whether the electrode is misaligned and adjust the welding parameters in time according to the output of the model to realize the automatic control of the welding process.

[0162] Refer to Figure 2 , an electrode misalignment detection system based on a bipolar Hall effect sensor, comprising:

[0163] The first module 201 is used to cancel the offset voltage and noise of the Hall effect sensor by the rotation current offset cancellation method and the chopper stabilization technology method to obtain a bipolar Hall effect sensor;

[0164] The second module 202 is used to collect and process the Hall voltage of the electrode based on a bipolar Hall effect sensor, and obtain the change amount of the contact area at the tip of the electrode;

[0165] The third module 203 is used to perform feature extraction processing on the change amount of the contact area at the tip of the electrode through a neural network model, determine the misalignment amount of the electrode, and compare it with the standard threshold misalignment amount to obtain the electrode misalignment detection result.

[0166] The content in the above method embodiments is applicable to the system embodiments of the present invention. The functions specifically implemented by the system embodiments of the present invention are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those of the above method embodiments.

[0167] The above is a specific description of the preferred embodiment of the present invention, but the present invention is not limited to the described embodiment. Those skilled in the art can also make various equivalent deformations or substitutions without departing from the spirit of the present invention, and these equivalent deformations or substitutions are all included within the scope defined by the claims of this application.

Claims

1. An electrode misalignment detection method based on a dual-pole Hall effect sensor, characterized in that Including the following steps: Eliminating the offset voltage and noise of the Hall effect sensor by the rotating current offset elimination method and the chopper stabilization technology elimination method to obtain a bipolar Hall effect sensor; Collecting and processing the Hall voltage of the electrode based on the bipolar Hall effect sensor to obtain the change amount of the contact area at the tip of the electrode; Performing feature extraction processing on the change amount of the contact area at the tip of the electrode through a neural network model, determining the misalignment amount of the electrode and comparing it with the standard threshold misalignment amount to obtain the electrode misalignment detection result.

2. The electrode misalignment detection method based on a dual-pole Hall effect sensor according to claim 1, wherein The bipolar Hall effect sensor includes a rotating current source, a Hall element, a modulation circuit, a chopper amplifier circuit, a demodulation circuit, a comparison circuit and an output control circuit, and the rotating current source, the Hall element, the modulation circuit, the chopper amplifier circuit, the demodulation circuit, the comparison circuit and the output control circuit are connected in sequence.

3. The electrode misalignment detection method based on a dual-magnetic-pole Hall effect sensor according to claim 2, wherein The step of eliminating the offset voltage and noise of the Hall effect sensor by the rotating current offset elimination method and the chopper stabilization technology elimination method to obtain a bipolar Hall effect sensor specifically includes: Based on the rotating current source, eliminating the offset voltage of the Hall effect sensor by the rotating current offset elimination method to obtain the Hall effect sensor after eliminating the offset voltage; Based on the chopper amplifier circuit, eliminating the offset voltage noise of the Hall effect sensor by the chopper stabilization technology elimination method to obtain the Hall effect sensor after eliminating the offset voltage noise; Combining the Hall effect sensor after eliminating the offset voltage and the Hall effect sensor after eliminating the offset voltage noise to obtain a bipolar Hall effect sensor.

4. The electrode misalignment detection method based on the bipolar Hall effect sensor according to claim 3, wherein The step of based on the rotating current source, eliminating the offset voltage of the Hall effect sensor by the rotating current offset elimination method to obtain the Hall effect sensor after eliminating the offset voltage specifically includes: Generating a vertical-phase Hall element bias voltage and a horizontal-phase Hall element bias voltage through the rotating current source; Based on the vertical-phase Hall element bias voltage, the supply current of the Hall element flows from top to bottom to obtain the vertical-phase output voltage of the Hall element; Based on the horizontal-phase Hall element bias voltage, the supply current of the Hall element flows from left to right to obtain the horizontal-phase output voltage of the Hall element; Superposing and averaging the vertical-phase output voltage and the horizontal-phase output voltage of the Hall element to obtain the Hall effect sensor after eliminating the offset voltage.

5. The electrode misalignment detection method based on a dual-pole Hall effect sensor according to claim 4, wherein The step of based on the chopper amplifier circuit, eliminating the offset voltage noise of the Hall effect sensor by the chopper stabilization technology elimination method to obtain the Hall effect sensor after eliminating the offset voltage noise specifically includes: Powering on the Hall effect sensor to obtain the Hall voltage to be processed; Based on the chopper amplifier circuit, generating a high-frequency square-wave modulation signal to modulate the Hall voltage to be processed to a high-frequency band to obtain a high-frequency modulation signal; Superposing and processing the high-frequency modulation signal with the offset voltage and noise of the Hall effect sensor to obtain a mixed signal; Amplifying the mixed signal and then demodulating it through the high-frequency square-wave modulation signal to obtain a baseband signal. Filter the high-frequency noise and offset voltage of the baseband signal to obtain a Hall effect sensor with the offset voltage noise eliminated.

6. The electrode misalignment detection method based on a bipolar Hall effect sensor according to claim 5, wherein The step of collecting the Hall voltage of the electrode by the bipolar Hall effect sensor and obtaining the change amount of the contact area at the tip of the electrode specifically includes: Apply an electric current to the electrode to obtain the current density of the electrode; According to Ampere's circuital law, determine the relationship between the magnetic field strength of the electrode and the current density of the electrode; Collect the Hall voltage of the electrode by the bipolar Hall effect sensor to obtain the Hall voltage of the electrode; Calculate the contact area of the electrode by combining the Hall voltage of the electrode and the relationship between the magnetic field strength of the electrode and the current density of the electrode to obtain the change amount of the contact area at the tip of the electrode.

7. The electrode misalignment detection method based on a dual-pole Hall effect sensor according to claim 6, characterized in that The expression for calculating the electrode contact area is specifically as follows: In the above formula, V H represents the Hall voltage of the electrode, K represents the sensitivity coefficient, k represents the proportionality coefficient, and C a represents the contact area of the electrode tip, and I represents the magnitude of the current passing through the electrode.

8. The electrode misalignment detection method based on a bipolar Hall effect sensor according to claim 7, characterized in that The step of extracting features from the change amount of the contact area at the tip of the electrode by the neural network model, determining the misalignment amount of the electrode, and comparing it with the standard threshold misalignment amount to obtain the electrode misalignment detection result specifically includes: Extract features from the change amount of the contact area at the tip of the electrode by the neural network model to obtain the mean and variance of the magnetic field strength; Determine the misalignment amount of the electrode according to the mean and variance of the magnetic field strength; Collect the Hall voltage of the standard electrode by the bipolar Hall effect sensor to obtain the misalignment amount of the standard electrode as the standard threshold misalignment amount; If the misalignment amount of the electrode is greater than the standard threshold misalignment amount, there is an axial misalignment problem with the electrode currently, and an audible and visual alarm is triggered; If the misalignment amount of the electrode meets the range of the standard threshold misalignment amount, there is no axial misalignment problem with the electrode currently, and the electrode misalignment detection result is output.

9. An electrode misalignment detection system based on a bipolar Hall effect sensor, characterized in that, It includes the following modules: The first module is used to eliminate the offset voltage and noise of the Hall effect sensor by the rotating current offset cancellation method and the chopper stabilization technique to obtain a bipolar Hall effect sensor; The second module is used to collect the Hall voltage of the electrode by the bipolar Hall effect sensor to obtain the change amount of the contact area at the tip of the electrode; The third module is used to extract features from the change amount of the contact area at the tip of the electrode by the neural network model, determine the misalignment amount of the electrode, and compare it with the standard threshold misalignment amount to obtain the electrode misalignment detection result.

Images