Laser light-transmitting scribing control method and system for solar thin-film cell

Abstract

The invention relates to the technical field of solar cell processing, in particular to a laser light-transmitting scribing control method and system for a solar thin-film cell. The control method comprises the steps that a large-size substrate is divided into a plurality of galvanometer view fields through a dynamic flight continuous machining mode, a continuous snake-shaped machining path is generated, the laser machining platform is driven to continuously move at the constant speed, and the galvanometer scanning system is synchronously triggered to start lineation in a splicing area in advance. According to the method, a time domain coupling equation of galvanometer scanning frequency-platform movement speed-laser pulse sequence is established, and dynamic compensation of an optical path and self-adaptive adjustment of energy density are achieved in the continuous machining process. Based on a view field expansion algorithm of a nonlinear kinematics model, the phase deviation of a scanning vector and a motion vector is effectively inhibited, and the geometric continuity of a splicing interface in ultra-long stroke machining is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of solar cell processing, and in particular to a method and system for controlling laser light transmission scribing of solar thin-film cells. Background Art

[0002] Lightweight thin-film solar modules are currently widely used in the photovoltaic building integration (BIPV) sector. However, their laser precision scribing process faces a bottleneck in multi-dimensional precision coordinated control. Traditional step-by-step processing is limited by the start-stop inertia of the motion platform and the static matching mechanism of the galvanometer field of view. In dynamic continuous processing scenarios, the system requires repeated mechanical positioning compensation and optical path recalibration, making it difficult to achieve spatiotemporal coupling between the galvanometer scanning trajectory and the platform motion vector. Particularly problematic is that when processing ultra-large, irregularly shaped solar cell substrates, existing galvanometer stitching algorithms fail to fully consider the real-time interaction between kinematic parameters and optical modulation factors. This results in phase delay accumulation and distortion of the energy density distribution along the scanning trajectory, ultimately forming periodic micro-defect bands at the stitching interface. These defects not only disrupt the electrical continuity of the transparent conductive layer but also induce non-uniform light field interference within the thin-film stack structure, limiting the optical transmittance of building photovoltaic components. Furthermore, the frequent calibration of the galvanometer field of view and platform movement in traditional stitching processes can lead to overlapping scribing lines or gaps greater than 10μm, forming MURA patterns that affect light transmission uniformity. Summary of the Invention

[0003] The present disclosure provides a method and system for controlling laser light transmission scribing of solar thin-film cells, aiming to overcome at least one defect in the prior art.

[0004] To achieve the above objectives, the technical solutions disclosed in the present invention are as follows: According to one aspect of the present disclosure, a method for controlling laser light transmission scribing of solar thin-film cells is provided, the control method comprising: Through the dynamic flight continuous processing mode, the large-scale substrate is divided into multiple galvanometer fields of view, and a continuous serpentine processing path is generated. The boundaries of adjacent galvanometer fields of view are formed into a splicing area, the laser processing platform is driven to move continuously at a constant speed, and the galvanometer scanning system is synchronously triggered to start scribing in advance in the splicing area; Small BOX format segmentation and splicing control is adopted to divide the entire processing area into multiple independent small BOX units. The range of the galvanometer field of view is dynamically expanded based on the real-time platform displacement. Through the closed-loop matching of encoder feedback and galvanometer scanning speed, the line width error is controlled within ±3μm. During the laser emission phase, the first pulse is suppressed, and ultrafast laser modulation technology is used to nonlinearly attenuate the peak power of the first N pulses, eliminating mura defects caused by abnormal first pulse energy. A transmittance feedback compensation mechanism is established based on the real-time coordinates of the flight path. The transmittance uniformity of the stitching area is synchronously scanned by a line array camera. When the transmittance difference between adjacent stitching areas exceeds a set threshold, the laser parameters and the marking interval are dynamically adjusted. The linked dust removal system follows the motion trajectory of the laser processing head, and the etching position is tracked in real time through an independent axially driven dust extraction device. A high-pressure gas injection module is integrated inside the dust extraction port to form a directional airflow barrier to suppress dust diffusion.

[0005] Furthermore, the steps of the dynamic flight continuous processing mode include: A dynamic expansion model of the galvanometer field of view is established. The expansion amount of the galvanometer field of view is calculated according to the platform movement speed and the maximum scanning angle of the galvanometer to generate an overlapping scanning area covering the adjacent field of view boundaries. The calculation formula for the expansion amount of the galvanometer field of view is: , where t0 is the initial time, t is the continuous processing time, v(t') is the instantaneous speed of the platform at time t', θ max is the maximum scanning angle of the galvanometer; Using a synchronous trigger algorithm, when the laser focus coordinates enter the critical area within 5% of the current field of view boundary, the scanning galvanometer deflection control of the next field of view is started; Energy gradient control is implemented in the overlapping scanning area, and the line width change rate of the overlapping scanning area satisfies the following conditions by coupling adjustment of the laser pulse frequency f and the duty cycle D: , where W is the line width; The synchronization triggering algorithm includes: The platform displacement is captured in real time by a high-precision encoder, and the vector coordinate sequence of the serpentine path is preloaded in the galvanometer scanning system; When the platform displacement reaches the field of view switching threshold, the galvanometer deflection controller is triggered to load the scanning parameters of the next field of view in advance. The loading time difference is less than 20% of the system response delay time, and the quadratic interpolation algorithm of the galvanometer scanning angle is used to eliminate the mechanical gap error during the loading process.

[0006] Furthermore, the steps of controlling the splitting and splicing of the small BOX formats include: A dynamic error compensation model is established based on the small BOX unit size. The influence of the platform motion acceleration a(t) on the splicing accuracy is eliminated by the galvanometer compensation amount. The calculation formula of the galvanometer compensation amount is: , where k is the galvanometer scanning resolution correction coefficient; A dual-weight fusion function is established at the boundaries of adjacent small boxes to perform smooth transition processing on the line coordinates: , where a represents the weight coefficient, x1 and x2 represent the original coordinates of the adjacent small BOXs, and the weight coefficients satisfy: , where d0 represents the length of the fusion interval.

[0007] Furthermore, the step of focusing on suppressing the first pulse includes: Establish the first pulse energy decay function: , where E n is the energy of the nth pulse, E0 is the initial pulse energy, γ is the attenuation coefficient, n is the pulse sequence number, and τ is the time constant; Apply nonlinear power modulation to the first five pulses to satisfy: , n∈[1,5], where P n is the power of the nth pulse, P max is the maximum power.

[0008] Furthermore, the transmittance feedback compensation mechanism includes: A transmittance balancing control algorithm was constructed, using a proportional-integral-differential control strategy to dynamically compensate for real-time transmittance deviations, where the integral term weight is inversely proportional to the platform's motion speed. The marking interval is nonlinearly adjusted according to the compensation amount. The adjustment function is a hyperbolic tangent nonlinear mapping relationship to ensure that the transmittance fluctuation amplitude does not exceed 15% of the initial value. When the platform speed increases to the preset threshold, the marking interval is automatically reduced to maintain transmittance uniformity.

[0009] Furthermore, the control logic of the dust removal following system includes: A dust extraction device position synchronization model was established, and the dust extraction port offset was dynamically adjusted according to the platform movement speed and the dust diffusion angle. The dust diffusion angle was calibrated in real time through particle trajectory analysis using a high-speed camera. A correlation control strategy between the high-pressure gas injection pressure and the platform speed is designed. The injection pressure increases linearly with the square of the speed, and a local negative pressure zone is formed at the front end of the dust extraction port to improve the dust capture efficiency.

[0010] Furthermore, the dynamic error compensation model includes: The galvanometer scanning angular velocity is pre-compensated according to the platform motion acceleration. An acceleration lag correction factor is introduced when calculating the compensation amount. The value of the correction factor is positively correlated with the torque response time of the platform drive motor. The compensated marking position is closed-loop verified by a laser interferometer.

[0011] Furthermore, the energy decay function includes: A pulse shaping module is set in the laser modulation circuit to apply nonlinear slew rate control to the rising edge of the first pulse. The slew rate adjustment amplitude has a piecewise linear relationship with the preset mura defect elimination level. The pulse energy is monitored in real time by a photodetector and fed back to the dynamic adjustment loop of the attenuation coefficient γ.

[0012] Furthermore, the step of adjusting the line spacing includes: A database of inversely proportional correlation between marking interval and platform speed is established. When the speed reaches a preset critical value, the interval compression mode is activated. The compression ratio is dynamically optimized based on the real-time transmittance sampling data. During the optimization process, a genetic algorithm is used to perform multi-objective optimization of the compression coefficient. The objective function includes the weighted sum of the transmittance uniformity score and the processing efficiency weight.

[0013] According to another aspect of the present disclosure, a solar thin-film cell laser light transmission scribing control system is provided, the control system comprising: A loading and unloading control module is used to fix the glass substrate through a lifting conveyor and a vacuum adsorption mechanism; Dynamic on-the-fly machining module, including a galvanometer scanning subsystem and a motion platform control unit, for achieving continuous serpentine path machining; Small BOX splicing module, used to execute dynamic field of view expansion and error compensation algorithm; Pulse suppression module, integrating pulse shaping circuit and energy monitoring feedback unit; Transmittance closed-loop control module, including a line array camera, an optical sensor, and a real-time parameter adjustment unit; Dust removal linkage module, with an axial drive mechanism and an air curtain generating device with adjustable injection pressure; A central controller is used to coordinate each module to execute the operation process of the above-mentioned solar thin-film cell laser light transmission scribing control method; The dynamic on-the-fly processing module further comprises: High-precision encoder, used to collect platform displacement in real time and convert it into galvanometer deflection control signal; Field of view stitching logic unit with built-in advance triggering algorithm to eliminate mechanical delay. The algorithm predicts the future displacement based on the platform acceleration and loads the galvanometer parameters 3-5 control cycles in advance. The energy gradient controller realizes line width uniformity control by adjusting the pulse frequency and duty cycle of the laser, and the control signal and the output of the transmittance feedback module form a cross-coupling compensation loop.

[0014] The beneficial effects of the present invention are: The present invention establishes a time-domain coupling equation of the galvanometer scanning frequency-platform motion speed-laser pulse sequence to achieve dynamic compensation of the optical path and adaptive adjustment of the energy density during continuous processing. The field of view expansion algorithm based on the nonlinear kinematic model enables the galvanometer scanning trajectory to match the platform displacement component in real time, effectively suppresses the phase deviation between the scanning vector and the motion vector, and ensures the geometric continuity of the splicing interface during ultra-long stroke processing. By introducing a pulse energy time-domain modulation mechanism, an energy gradient transition is implemented at the turning point of the scanning path to eliminate the thermal stress mutation caused by traditional step-type power regulation. Combined with the gas-solid coupling flow field optimization design, a localized suction-purge collaborative system is constructed to form a laminar covering layer at the dynamic processing interface, which simultaneously solves the problems of processing debris removal and heat-affected zone control, and improves the optical uniformity of the transparent conductive layer and the bonding strength of the heterogeneous interface.

[0015] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention and implement it according to the contents of the specification, the following is a detailed description of the preferred embodiments of the present invention with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 This is a flow chart of a method for controlling laser light transmission scribing of solar thin-film cells according to one embodiment of the present invention; Figure 2 Schematic diagram of the modified galvanometer field of view expansion model in one embodiment of the present invention; Figure 3 A schematic diagram of a dynamic error compensation vector field according to an embodiment of the present invention; Figure 4 Schematic diagram of the first pulse energy attenuation curve in one embodiment of the present invention; Figure 5 This is a heat map of light transmittance distribution in one embodiment of the present invention; Figure 6 A schematic diagram of a dust removal airflow field simulation in one embodiment of the present invention; Figure 7 Schematic diagram of a line spacing adjustment curve in one embodiment of the present invention. DETAILED DESCRIPTION

[0017] The technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the present invention.

[0018] The present invention provides the following preferred embodiments: Example 1 Aiming at the problem of interface transmittance distortion caused by the spatiotemporal mismatch between the scanning trajectory of the galvanometer and the motion vector of the platform during the dynamic continuous processing of large-scale solar thin-film cells, this embodiment proposes a control method for laser light transmission scribing of solar thin-film cells, which optimizes the coordinated control mechanism of the galvanometer field of view expansion and kinematic parameters in the on-the-fly continuous processing mode. Figure 1 As shown, the steps of the laser light transmission scribing control method include: S100: Through the dynamic flight continuous processing mode, the large-size substrate is divided into multiple galvanometer fields of view, generating a continuous serpentine processing path. The boundaries of adjacent galvanometer fields of view form a stitching area, driving the laser processing platform to move continuously at a constant speed, and synchronously triggering the galvanometer scanning system to start scribing in the stitching area in advance.

[0019] S200: It uses small BOX format segmentation and splicing control to divide the entire processing area into multiple independent small BOX units. The galvanometer field of view is dynamically expanded based on real-time platform displacement. Through closed-loop matching of encoder feedback and galvanometer scanning speed, the marking width error is controlled within ±3μm.

[0020] S300: Implements first-pulse focused suppression during the laser emission phase, using ultrafast laser modulation technology to nonlinearly attenuate the peak power of the first N pulses to eliminate mura defects caused by abnormal first-pulse energy. S400: A transmittance feedback compensation mechanism is established based on the real-time coordinates of the flight path. The line array camera is used to synchronously scan the transmittance uniformity of the stitching area. When the transmittance difference between adjacent stitching areas exceeds the set threshold, the laser parameters and marking interval are dynamically adjusted. S500: The dust removal system is linked to the motion trajectory of the laser processing head. The etching position is tracked in real time through an independent axially driven dust extraction device, and a high-pressure gas injection module is integrated inside the dust extraction port to form a directional airflow barrier to suppress dust diffusion.

[0021] As the platform moves continuously at a constant speed, the galvanometer system dynamically adjusts the scanning angle based on the real-time displacement feedback from the encoder, causing a gradient shift in the laser beam's angle of incidence in the adjacent field of view stitching area, forming a continuous serpentine scanning trajectory. It is important to understand that this field of view expansion model achieves precise matching of the galvanometer scanning speed and the platform's motion speed in the time domain by integrating the historical time relationship between the platform velocity and the maximum scanning angle, ensuring that the optical path error during dynamic processing converges to the submicron level.

[0022] Furthermore, this embodiment refines the small BOX format segmentation strategy. Specifically, the entire processing area is divided into multiple independent units, each of which is slightly larger than the nominal field of view of the galvanometer. Within a single unit, the galvanometer scanning speed is segmented and adjusted based on the real-time displacement of the platform: when the platform runs at a constant speed, the galvanometer speed maintains a fixed ratio with the platform; when the platform is accelerated by a disturbance, the galvanometer speed is adjusted by the platform. Figure 3 The vector field model shown calculates the compensation in real time and adds it to the galvanometer deflection command. It can be understood that this compensation algorithm significantly reduces the phase jump error at the border between adjacent cells through encoder feedback closed-loop control.

[0023] To address the interface defects caused by abnormal first pulse energy, this embodiment uses ultrafast laser modulation technology. Figure 4 As shown in the figure, during the initial laser emission phase, the peak power of the first N pulses is nonlinearly attenuated, causing the energy to gradually transition to a steady-state value according to an exponential law. It should be noted that this modulation generates real-time control signals through hardware circuits to suppress the overshoot of the first pulse caused by the charging and discharging delays of the energy storage element, while also avoiding sudden thermal stress changes through a gradual transition zone.

[0024] At the same time, this embodiment constructs an online feedback compensation system for transmittance. Figure 5 As shown, a line scan camera collects a real-time heat map of the transmittance distribution in the stitching area. When the transmittance difference between adjacent areas exceeds a threshold, the laser parameters and the scribing interval are dynamically adjusted. This adjustment mechanism, through a speed-spacing coupling function, automatically reduces the spacing at high speeds to compensate for thermal diffusion and increases the spacing at low speeds to prevent excessive ablation, thereby maintaining global uniformity.

[0025] The benefit of this embodiment is that it realizes multi-dimensional parameter collaborative optimization by integrating dynamic field of view expansion, error compensation algorithm and transmittance feedback mechanism. Figure 6 As shown, the dust removal system utilizes an independent axial drive mechanism, whose motion trajectory is synchronized with the laser processing head. A high-pressure jet module is integrated within the dust extraction port to form a directional laminar flow barrier. This design, optimized through airflow field simulation, forces dust particles to converge toward the dust extraction port in a vortex field, avoiding the disturbance of the processing area caused by traditional extraction. This embodiment achieves millimeter-level tracking accuracy in the dust removal area without adding additional motion axes, significantly reducing the risk of contamination.

[0026] Example 2 In order to solve the problem of uneven marking caused by lag in field of view switching and energy mutation in dynamic flight continuous processing, this embodiment refines the field of view dynamic expansion model and synchronous triggering mechanism, and constructs a coupling adjustment method for energy gradient control. Figure 2As shown, the dynamic expansion model of the galvanometer field of view calculates the field of view expansion ΔL in real time by analyzing the temporal and spatial relationship between the platform motion speed and the galvanometer scanning angle. Specifically, the expansion amount is calculated as follows: , where t0 is the initial time, t is the continuous processing time, v(t') is the instantaneous speed of the platform at time t', θ max is the maximum scanning angle of the galvanometer, and its value is calibrated by the deflection range of the galvanometer motor and the focal length of the f-theta lens. It should be understood that the integral model dynamically associates the velocity history data with the tangent value of the scanning angle to generate an overlapping area covering the adjacent field of view boundaries, such as Figure 2 As shown in the shaded part, its width is adaptively adjusted as the platform speed changes.

[0027] Furthermore, this embodiment optimizes the synchronous trigger logic of field of view switching. The serpentine path vector coordinate sequence optimized by the Bézier curve is preloaded in the galvanometer control system, and the data point interval is set to 5μm. When the high-precision magnetic grating encoder detects that the platform displacement reaches 95% of the current field of view boundary, the galvanometer is automatically triggered. Figure 4 The displacement-time prediction curve shown triggers the pre-loading of the scanning parameters for the next field of view. It can be understood that the trigger algorithm adopts a double buffer architecture, and the control instructions of the current field of view and the next field of view are processed in parallel in the galvanometer deflection controller, so that the parameter loading time difference is shortened to within 18% of the system response delay time. At the same time, during the switching process, the galvanometer deflection angle is interpolated with a quadratic polynomial, such as Figure 4 As shown in the smooth transition section of the mid-angle transition curve, the 0.6-1.1μm positioning error caused by the gear backlash is effectively compensated.

[0028] This embodiment specifically designs an energy gradient control module for the overlapping area. Figure 5 As shown in the figure, when the laser focus enters the critical area within the first 5% of the adjacent field of view boundary, the differential relationship between the platform instantaneous velocity v(t) and the line spacing is calculated in real time by the FPGA, and the coupling parameters of the pulse frequency f and the duty cycle D are dynamically adjusted. Specifically, the line width change rate constraint condition is: , where W represents the width of the line and x represents the coordinate position along the line path. When the platform speed is increased to 300 mm / s, the pulse frequency is linearly increased from 180 kHz to 320 kHz, while the duty cycle is reduced from 35% to 20% in a piecewise function, so that the energy density per unit length remains constant. It should be noted that this adjustment process is achieved by Figure 5 The exponential decay curve shown performs a gradual energy transition on the first and last pulses in the overlapping area, achieving seamless energy connection within the overlapping band with a width of 1.2-2.0 mm.

[0029] Furthermore, this embodiment improves the real-time performance of the data processing link. A high-speed integrator module is embedded in the FPGA, and a sliding time window mechanism is used to perform weighted integration on the platform velocity history data. The window length is set to 400ms to balance the calculation accuracy and real-time requirements. At the same time, a galvanometer scanning angle-position mapping table is pre-stored in the DSP, and the integration result is converted into a galvanometer motor deflection instruction through a table lookup method, so that the field of view expansion calculation cycle is compressed to 0.1ms. It can be understood that the above design makes the calculation error of the dynamic expansion model stable within the range of ±0.7μm, such as Figure 2 The error distribution curve is shown.

[0030] This embodiment also enhances the coordinated control capability of the dust removal system. Figure 6 As shown, a multi-stage Venturi structure is integrated at the dust extraction nozzle, and the air flow diffusion angle is optimized to 58 degrees through computational fluid dynamics simulation. When the platform movement speed changes, the dust removal module Figure 4 The velocity-flow mapping curve in the image is used to adjust the negative pressure value synchronously, so that the dust capture area always covers the range of 2-4mm behind the laser processing point. Figure 6 The laminar flow effect of the midstream distribution increases the capture efficiency of particles with a diameter greater than 3μm to more than 97%, while avoiding the line trajectory disturbance caused by traditional lateral suction elicitation.

[0031] The benefit of this embodiment is that it accurately predicts the field of view expansion through the integral model, and achieves seamless connection of high-speed continuous processing by combining synchronous triggering and energy gradient control. Figure 2 and Figure 5 As shown by the synergistic effect of the dynamic expansion model, the field of view switching hysteresis error is suppressed to the order of ±0.5μm, while the energy gradient algorithm makes the line width uniformity of the overlapping area reach the submicron level and smooth transition. Figure 6 The optimized fluid design shown effectively maintains the cleanliness of the processing area and reduces the optical performance degradation caused by dust redeposition on the film surface.

[0032] Example 3 In order to solve the problems of splicing misalignment and boundary mutation caused by platform acceleration disturbance in small BOX format segmentation, this embodiment constructs a dynamic correction model of the galvanometer based on kinematic compensation, and innovatively proposes a dual-weight fusion algorithm to optimize the quality of the splicing transition area. Figure 3 As shown, the dynamic error compensation model calculates the galvanometer compensation amount Δx(t) in real time by analyzing the spatiotemporal coupling effect of the platform motion acceleration on the galvanometer scanning path. Specifically, the compensation amount calculation formula is: , where k is the galvanometer scanning resolution correction coefficient, and its value is determined by the microstep subdivision parameters of the galvanometer stepper motor and the magnification of the optical system; v(t) is the instantaneous velocity of the platform at time t, which is collected in real time by the grating encoder with a time resolution of 0.1μs; a(t) is the platform acceleration, which is solved by the fusion of the acceleration sensor and the velocity differential algorithm; Δt is the system response delay time, covering the full link time lag from encoder signal acquisition to galvanometer deflection command output. It should be understood that this model can effectively offset the galvanometer tracking lag caused by the inertial force in the acceleration and deceleration stages by introducing the platform acceleration into the quadratic compensation, such as Figure 3 The compensated gradient of the vector field distribution is shown in Figure 5.

[0033] Furthermore, this embodiment implements a dual-weight coordinate fusion strategy at the splicing boundary of adjacent small BOX units. Figure 7 As shown in the figure, when the laser focus enters the boundary area, the coordinate data of the adjacent cells are smoothly transitioned through the following function: , where α is the weight coefficient, x1 and x2 are the original coordinates of the adjacent small boxes. The weight coefficient is dynamically generated by the Sigmoid function: , where d represents the normalized distance from the current focus to the boundary, d0 is the length of the fusion interval, which is set to 0.5mm to cover the heat-affected zone; β is the transition slope adjustment factor, whose value is adaptively adjusted according to the thermal diffusion coefficient of the film material. For example, for amorphous silicon film, β=12, while for microcrystalline silicon film, β=8. It can be understood that this function converts the coordinate jump of the splicing boundary into a continuous gradient through an exponential smooth transition, such as Figure 7 The smooth connection at the inflection point of the middle curve is shown.

[0034] This embodiment also optimizes the data processing link to improve real-time performance. A galvanometer compensation calculation module is preset in the FPGA, and a parallel pipeline architecture is used to synchronously process the acceleration and velocity signals to ensure that the compensation instruction update frequency reaches 10kHz. At the same time, a fusion function fast lookup algorithm is embedded in the DSP, and the coordinate fusion calculation time is shortened to less than 5μs by precalculating the α distribution curve corresponding to different β values. It should be noted that the above hardware collaborative design enables the system to achieve a maximum platform acceleration of 200mm / s 2 Under working conditions, the galvanometer compensation response time can still be guaranteed to be less than 1ms.

[0035] Furthermore, this embodiment introduces a feedback mechanism for the transmittance of the boundary area. Figure 5As shown, the transmittance distribution of the fusion area is monitored in real time using a linear CCD array. When the transmittance gradient exceeds a preset threshold, the values of d0 and β are dynamically adjusted. For example, when the transmittance fluctuation exceeds 3%, d0 is extended to 0.8mm and β is reduced to 6 to extend the fusion range and slow the transition slope, thereby suppressing microcracks caused by differences in the thermal expansion coefficients of the materials.

[0036] The benefit of this embodiment is that it accurately offsets acceleration disturbances through the kinematic compensation model and achieves nanometer-level splicing accuracy in combination with the dual-weight fusion algorithm. Figure 3 and Figure 7 As shown by the synergistic effect of the compensation model, the galvanometer tracking error during the platform acceleration and deceleration phase is suppressed to within ±0.4μm, while the fusion algorithm reduces the line width change rate of the splicing boundary to below 0.05μm / mm. Figure 5 As shown in the transmittance heat map distribution, the feedback regulation mechanism ensures that the standard deviation of transmittance uniformity in the splicing area of different material systems is less than 1.2%, which improves the electrical performance consistency of large-size thin-film batteries.

[0037] Example 4 To address the problem of thin film thermal damage caused by the sudden change in the first pulse energy at the initial stage of laser processing, this embodiment constructs a first pulse energy attenuation function and a nonlinear power modulation strategy, and combines timing synchronization optimization to achieve precise control of the heat-affected zone. Figure 4 As shown, the first pulse energy decay function adopts an exponential decay model, and its mathematical expression is: , where E n is the energy of the nth pulse, E0 is the initial pulse energy reference value, which is calibrated by the laser output power and the Q switch delay parameter; γ is the attenuation coefficient, ranging from 0.3 to 0.6, and is dynamically adjusted according to the extinction coefficient of the thin film material; n is the pulse sequence number, counting from 1; τ is the time constant, whose value is determined by the laser repetition frequency and the thermal relaxation time of the material. It should be understood that this function reduces the energy gradient to within 15% of the steady-state value within 5 cycles after the first pulse through the rapid decay characteristics of the exponential term, such as Figure 4 As shown in the smooth transition section of the energy envelope curve.

[0038] Furthermore, this embodiment implements nonlinear power modulation constraints on the first five pulses: , n∈[1,5], where P n is the power of the nth pulse, P max Specifically, at the start-up stage of the pulse sequence, the rising edge slope control function of the acousto-optic modulator (AOM) is used to limit the first pulse power to 0.45P maxThe power then increases to the steady state power according to the square root rule. It can be understood that this constraint condition reduces the edge melting phenomenon caused by the heat accumulation effect by suppressing the power jump rate. Figure 5 As shown in the figure, the transmittance distribution thermogram shows that the optimized first pulse modulation reduces the standard deviation of the temperature distribution in the heat-affected zone to less than 31% of that before optimization.

[0039] Furthermore, this embodiment integrates a dynamic waveform generation module into the laser control system. The DSP chip is embedded with a piecewise linear approximation algorithm, which discretizes the energy attenuation function into a 256-level adjustable amplitude sequence and outputs it to the laser drive circuit through a 16-bit digital-to-analog converter. It should be noted that the waveform generator synchronously receives the real-time energy feedback signal from the photodiode. When it detects that the energy deviation of the first pulse exceeds ±5%, it automatically triggers the PID adjustment algorithm to correct the AOM drive voltage to ensure E n The actual output error is less than ±2.5%.

[0040] Furthermore, this embodiment designs a material adaptive parameter matching mechanism. For different film types, the optimal combination of γ and τ is calibrated through experiments: for amorphous silicon film, γ=0.55 and τ=2.5 are set; for CIGS film, γ=0.45 and τ=1.6 are used. The above parameters are stored in the FPGA lookup table, and the dynamic switching of processing parameters is achieved through the host computer instructions. Figure 7 As shown in the figure, when the platform speed is increased to 400 mm / s, the marking interval is adaptively adjusted according to the hyperbolic tangent function, and its changing gradient forms a synergistic control with the pulse energy modulation to ensure the balance between processing efficiency and quality.

[0041] This embodiment also optimizes the spatiotemporal synchronization of pulse triggering and platform motion. The laser Q switch trigger signal and the galvanometer deflection instruction are aligned through the time-to-digital converter (TDC), so that the position error between the first pulse emission moment and the galvanometer scanning starting point is less than 0.3μm. It should be understood that this synchronization mechanism is based on the delay locked loop (DLL) in the FPGA, and its clock resolution reaches 12ps, which effectively eliminates the misalignment of the starting point of the marking caused by timing jitter. Figure 3 As shown in FIG, the dynamic error compensation vector field generates a gradient distribution of the galvanometer compensation amount by jointly modeling the velocity and acceleration, thereby further suppressing the positioning deviation of the starting point.

[0042] The benefit of this embodiment is that the thermal shock of the first pulse on the film is significantly reduced through the synergistic effect of the energy attenuation function and the power constraint. Figure 4 and Figure 5 As shown in the comparative analysis, the smooth transition of the energy envelope reduces the lattice defect density in the starting area of the scribing to less than 28% of the traditional method, and the transmittance uniformity is improved to more than 97.5%. Figure 7The dynamic adjustment mechanism of the scribing interval shown achieves the unity of efficiency and precision under high-speed processing conditions, providing a reliable guarantee for the continuous processing of complex patterns.

[0043] Example 5 To address the issue of reduced photoelectric conversion efficiency for thin-film solar cells caused by uneven transmittance distribution, this embodiment optimizes the transmittance feedback compensation mechanism, achieving dynamic balancing of transmittance distribution through the synergistic effect of a closed-loop control algorithm and nonlinear interval adjustment. Specifically, during laser processing, a fiber optic spectrometer integrated into the side of the scanner monitors transmittance data of the scribed area in real time at a sampling frequency of 100Hz. Its detection wavelength range is 400-1100nm, covering the absorption spectrum characteristics of the thin-film material. The transfer function of the proportional-integral-derivative (PID) control strategy is expressed as: , where e(t) is the deviation between the set and measured transmittance values, v(t) is the platform's real-time motion velocity, and α is the velocity coupling coefficient, set to 0.02s / mm to suppress integral saturation during high-speed motion. It is important to understand that with the velocity term introduced into the denominator of the integral term, when the platform speed exceeds 200mm / s, the integral action strength automatically decays to below 60% of the low-speed state, thereby avoiding transmittance oscillations caused by overcompensation.

[0044] Furthermore, the compensation amount is mapped to the line spacing adjustment amount through the hyperbolic tangent function, and its mathematical expression is: , where s0 is the reference interval (50 μm), β is the adjustment gain coefficient (the value is 3.5), and e max is the maximum allowable transmittance deviation (15%). This function uses nonlinear transformation to limit the compensation to a range of ±0.2s0, ensuring a smooth transition of interval adjustment. At the hardware implementation level, the FPGA's embedded lookup table discretizes the transmittance deviation into 256 quantized values. This is output to the galvanometer control circuit via a 16-bit DAC, adjusting the frequency to synchronize with the laser pulse.

[0045] This embodiment also designs a critical trigger mechanism for speed adaptation. When the platform speed increases to a preset threshold (400 mm / s), the interval compression mode is automatically activated, and the compression ratio is determined by the following formula: , where v c is the critical speed, and k is the slope factor (0.008s / mm). This logic function makes the interval compression increase smoothly with the speed overshoot, avoiding the transmittance jump caused by sudden changes. It should be noted that after the compression mode is started, the proportional coefficient K of the PID controller p A simultaneous 30% improvement is made to enhance the system's dynamic response capabilities to high-speed conditions.

[0046] The benefits of this embodiment lie in achieving highly stable control of transmittance distribution during machining through the synergistic combination of closed-loop feedback of transmittance deviation and nonlinear interval adjustment. The inverse proportional relationship between the integral term and velocity effectively balances the conflict between dynamic response and steady-state accuracy, while the hyperbolic tangent mapping mechanism ensures smoothness and controllability of the adjustment process. Furthermore, the critical velocity-triggered interval compression mode provides adaptive process parameter optimization capabilities for high-speed machining.

[0047] Example 6 To address the issue of transmittance degradation caused by dust accumulation during laser processing, this embodiment refines the control logic of the dust removal tracking system and improves dust capture efficiency through a coupled design of spatial synchronization and pressure regulation. Specifically, the dust extraction device is driven by a six-degree-of-freedom robotic arm, the end of which is equipped with a circular array of dust extraction ports with an inner diameter of 8mm and an outer diameter of 12mm. The relative position of the port to the laser head is calibrated in real time using a high-precision encoder. The offset calculation formula of the position synchronization model is: , where v is the platform's instantaneous velocity, a is its acceleration, Δt is the system response delay (50ms), θ is the dust diffusion angle (obtained through Hough transform analysis of particle trajectories captured by a high-speed CMOS camera), and r is the diffusion radius (positively correlated with machining power). It's important to understand that this model uses a feedforward compensation mechanism to predict the trajectory of the dust cloud, keeping the position deviation of the dust extraction port within ±0.5mm.

[0048] Furthermore, the high-pressure gas injection pressure is dynamically adjusted according to the platform speed, and its control law is: , where P0 is the basic pressure (0.2 MPa), λ is the pressure coefficient (5×10 -6 MPa·s² / mm²). When the speed reaches 500 mm / s, the injection pressure rises to 0.325 MPa, forming a local negative pressure zone 20 mm in front of the dust extraction port. The pressure gradient is optimized to -150 Pa / mm through computational fluid dynamics (CFD) simulation. In specific implementation, the proportional valve receives the squared velocity signal from the motion controller and uses PID control to control the actual pressure error to within ±2%.

[0049] This embodiment also integrates a dust concentration feedback correction module. A light scattering particle sensor installed in the dust extraction duct monitors the dust concentration at a frequency of 1kHz. When the concentration exceeds the threshold, the following correction actions are automatically triggered: 1) A sinusoidal disturbance component (amplitude ±0.3 mm, frequency 10 Hz) was added to the robot arm's motion trajectory to expand the capture range; 2) The injection pressure is increased by 10% to enhance the intensity of the negative pressure field.

[0050] It should be noted that the correction parameters are stored in the EEPROM and the preset scheme can be called according to the type of processing material through the host computer interface.

[0051] The benefits of this embodiment lie in achieving dynamic matching between dust capture efficiency and processing speed through the coordinated optimization of the kinematic model and pressure control. Feedforward compensation for the dust extraction port position effectively overcomes tracking errors caused by system latency, while the velocity-squared pressure regulation strategy ensures an adaptive balance between negative pressure field intensity and dust diffusion rate. Furthermore, a feedback correction mechanism based on real-time concentration monitoring further enhances the system's adaptability to complex operating conditions.

[0052] Example 7 To address the problem of marking position deviation caused by inertial lag during the platform acceleration and deceleration phase, this embodiment optimizes the dynamic error compensation model and improves positioning accuracy by combining the pre-compensation algorithm with closed-loop verification. The pre-compensation amount of the galvanometer scanning angular velocity is determined by the following formula: , where ω0 is the nominal angular velocity, a is the instantaneous acceleration of the platform, τ m is the motor torque response time constant (calibrated to 12ms through step response test), J is the galvanometer moment of inertia (0.002kg·m²), τ c is the hysteresis correction factor (the value is 1.5τ m ). It should be understood that this formula simulates the inertia delay characteristics of the drive system through the exponential term, so that the compensation amount asymptotically approaches the theoretical demand value over time.

[0053] The compensated marking position is then verified in a closed-loop process using a laser interferometer. The interferometer's measurement beam is coaxial with the machining laser and detects the position offset of the galvanometer reflector at a sampling frequency of 200kHz with a resolution of 0.1μm. The verification algorithm uses a sliding window root mean square error (RMS) calculation. When the RMS error of five consecutive pulses exceeds 2μm, the following correction process is automatically triggered: 1) Update τ in the compensation model m parameter; 2) Recalibrate the moment of inertia J of the galvanometer; 3) Adjust the decay rate τ of the exponential term c It should be noted that the correction parameters are iteratively optimized by fusing historical data and real-time measurement values through the Kalman filter.

[0054] This embodiment also designs an acceleration partition compensation strategy. According to the sign and magnitude of the platform acceleration, the compensation amount is divided into three intervals: 1) When |a|<50mm / s 2 When , a linear compensation model is used; 2) When 50mm / s 2 ≤∣a∣<150mm / s 2 When , the enhanced compensation including the quadratic term is enabled; 3) When |a|≥150mm / s 2 When the threshold is calibrated experimentally, the thresholds for each interval are stored in a lookup table in the FPGA to enable fast switching.

[0055] The benefits of this embodiment include significantly reducing marking position deviations caused by sudden acceleration changes through modeling of inertial hysteresis characteristics and real-time closed-loop verification. The partitioned compensation strategy accounts for the dynamic characteristics differences between different acceleration ranges, while the high-precision measurement of the laser interferometer provides a reliable data foundation for the continuous optimization of the compensation model. Furthermore, the introduction of the Kalman filter effectively improves the noise resistance of the parameter correction process.

[0056] Example 8 To eliminate mura defects caused by sudden energy changes in the first laser pulse, this embodiment improves the implementation of the energy decay function, optimizing energy output characteristics through the coordinated control of pulse shaping and dynamic feedback. The pulse shaping module is integrated into the laser driver circuit and uses a field-programmable gate array (FPGA) to control the gate slew rate of the MOSFET switch. The rising edge slew rate adjustment function is: , where S max is the maximum slew rate (100V / ns), γ d is the attenuation coefficient corresponding to the mura defect elimination level (0.3-0.6), L is the real-time monitored pulse energy, L0 is the set energy value, and ΔL is the hysteresis range (5%). It is important to understand that this function uses a slew rate curve to achieve smooth switching, avoiding electromagnetic interference caused by step changes.

[0057] Furthermore, the dynamic adjustment loop of the attenuation coefficient γ uses a photodiode array for energy monitoring. After each pulse, the ADC collects light intensity data with 12-bit resolution and fits the energy attenuation curve using the least squares method. The residual is used to update the γ value: , where μ is the learning rate (0.01), and the summation term is the difference between the measured and modeled energies for the first five pulses. When the residual root mean square error for three consecutive pulses is less than 2%, the γ value is considered converged and the adjustment loop is locked. It should be noted that this algorithm uses iterative learning to compensate for characteristic drift caused by laser aging.

[0058] This embodiment also designs a segmented mapping relationship between Mura level and slew rate. According to the scoring results of the defect detection camera, the elimination level is divided into 4 levels: Level 1 corresponds to γ d=0.3 and Sr=70V / ns; Level 2 corresponds to γ d =0.4 and Sr=60V / ns; Level 3 corresponds to γ d =0.5 and Sr=50V / ns; Level 4 corresponds to γ d =0.6 and Sr=40V / ns, where Sr is the slew rate. Parameters for each level are optimized through orthogonal experiments and stored in non-volatile memory.

[0059] The benefits of this embodiment include: precise slew rate control and closed-loop energy feedback adjustment effectively suppress thin film defects caused by sudden changes in the first pulse energy. The segmented mapping relationship provides differentiated process parameters for different levels of mura reduction, while the iterative learning algorithm ensures the energy attenuation model's adaptability to changes in device state. Furthermore, the introduction of a sigmoid function prevents parameter oscillation during the adjustment process.

[0060] Example 9 In order to achieve adaptive matching between the marking interval and the processing speed, this embodiment optimizes the interval adjustment mechanism, and balances the transmittance and processing efficiency by combining database association with a multi-objective optimization algorithm. The inverse proportional association database uses an SQLite embedded structure to store a mapping table of platform speed-interval reference values. Its data points are calibrated by the response surface experiment method, covering a speed range of 0-1000mm / s and an interval resolution of 1μm. When the speed exceeds the critical value v c (400mm / s), the interval compression is determined by the following formula: , where T avg is the average transmittance of the most recent 50 sampling points, and η is the compression factor, obtained through genetic algorithm optimization. It is important to understand that this formula linearly maps normalized transmittance data to compression, ensuring that the adjustment process is linked to the process status in real time.

[0061] Furthermore, the objective function of the genetic algorithm is defined as: , where U is the transmittance uniformity score (based on the inverse of the standard deviation of the gridded sampling points), t total is the processing time per piece, w1 and w2 are weight coefficients (default settings are 0.7 and 0.3). The algorithm uses real number encoding, a population size of 100, a crossover probability of 0.85, a mutation probability of 0.01, and an iteration termination criterion of less than 1% fitness change for 20 consecutive generations. It should be noted that the optimization results are stored in a database and used as initial parameters for subsequent processing, shortening online calculation time.

[0062] This embodiment also designs a dynamic sensing module for the processing status. The force sensor integrated in the galvanometer system monitors the vibration acceleration of the laser focusing mirror in real time. When the energy of the vibration spectrum in the range of 200-500Hz exceeds the threshold, the interval compensation is automatically triggered: 1) Calculate the maximum allowable processing speed based on the main vibration frequency; 2) Reduce the spacing compression proportionally to prevent chatter marks.

[0063] It should be understood that this mechanism implements fast Fourier transform (FFT) through a hardware accelerator to ensure the real-time nature of the compensation response.

[0064] The benefits of this embodiment lie in the intelligent adjustment of marking intervals achieved through the collaboration of the database and optimization algorithm. The genetic algorithm's multi-objective optimization framework effectively balances the conflicting demands of transmittance uniformity and processing efficiency, while the vibration sensing and compensation mechanism enhances the system's adaptability to mechanical disturbances. Furthermore, the inversely proportional correlation database provides the foundation for rapid parameter retrieval for high-speed machining.

[0065] Example 10 In order to solve the problem that the response delay of the galvanometer and the uniformity of the line width are difficult to coordinately control in dynamic flight processing, this embodiment refines a laser processing system that integrates high-precision displacement feedback and cross-coupling compensation. Specifically, the high-precision encoder in the dynamic flight processing module adopts an absolute grating structure with a resolution of 0.1μm, and the output signal is transmitted to the interpolation operation unit of the galvanometer controller via the RS485 interface. The encoder collects the platform displacement in real time and generates the instruction value of the galvanometer deflection angle through linear transformation, where the angle conversion coefficient is calibrated based on the geometric relationship between the laser focus position and the curvature radius of the galvanometer mirror surface. It should be understood that the parallelism error between the installation position of the encoder and the platform guide rail is controlled within ±5μm / m to avoid the influence of Abbe error on the position feedback accuracy.

[0066] Furthermore, the field of view stitching logic unit has a built-in Kalman prediction algorithm, whose state equation is constructed based on the real-time data of the platform accelerometer. The prediction model updates the platform's displacement within the next three control cycles (i.e., 15ms) with a period of 5ms, and loads the corresponding galvanometer deflection parameters in advance. In the specific implementation, the platform acceleration signal is passed through a second-order Butterworth filter to eliminate high-frequency noise, and then input into the displacement predictor to generate a feedforward compensation amount. It can be understood that the algorithm dynamically adjusts the trigger time advance of the galvanometer command to a linear function of the platform speed. When the speed exceeds 300mm / s, the advance amount increases from 3 cycles to 5 cycles to offset the lag effect of the mechanical transmission chain.

[0067] The energy gradient controller is integrated into the FPGA of the laser driver board, and its pulse frequency adjustment range is 1-500kHz, and the duty cycle adjustment accuracy is 0.1%. The controller receives the compensation signal output by the transmittance closed-loop control module, and maps the transmittance deviation to the pulse parameter adjustment amount through the table lookup method. It should be noted that the adjustment direction of the pulse frequency and the duty cycle is opposite: when the transmittance is lower than the set value, the frequency is increased and the duty cycle is reduced to increase the number of laser actions per unit area; otherwise, the frequency is reduced and the duty cycle is increased to reduce the heat-affected zone. This cross-coupling strategy optimizes the weight coefficient through the orthogonal experimental method to ensure that the line width variation is controlled within ±2μm.

[0068] Furthermore, the small BOX splicing module uses a piezoelectric ceramic micro-motion stage to drive the optical beam expander group, with an adjustment range of ±1.5mm and a step accuracy of 0.05μm. During the field of view expansion process, the micro-motion stage dynamically adjusts the beam expander spacing according to the platform position signal, so that the effective processing field of view is expanded from the standard 30mm×30mm to 45mm×45mm. It can be understood that the displacement of the beam expander group is obtained by orthogonal decomposition of the platform movement direction, and a cosine gradient transition zone is superimposed at the boundary of the field of view to avoid splicing marks. In addition, the error compensation algorithm fine-tunes the beam expander position based on the transmittance detection results of the overlapping areas of adjacent fields of view, and the correction amount is generated by least squares fitting.

[0069] The pulse suppression module includes an IGBT-based pulse shaping circuit, whose rise time can be continuously adjusted between 5 and 50 ns. The energy monitoring feedback unit adopts a dual-path differential detection structure. The main light path is focused on the processing area, and the reference light path is introduced into the integrating sphere energy meter through a spectroscope. After the two signals are processed by the logarithmic amplifier, an energy attenuation compensation coefficient is generated and fed back to the IGBT drive circuit. It should be noted that the module inserts 5 gradually increasing pulses during the laser startup phase, and its energy is stepped up from 30% of the rated value to 100% to eliminate the instability of the first pulse.

[0070] In the transmittance closed-loop control module, a line scan camera and optical sensor are positioned at a 30° angle, with the line scan camera's scanning frequency triggered synchronously with the laser pulses. The optical sensor utilizes a four-quadrant detector structure, and its output signal is normalized and fed into a fuzzy PID controller, generating adjustment commands for laser power and focusing lens position. As can be appreciated, this module uses a timestamp alignment mechanism to correlate transmittance measurement data with platform position information, ensuring temporal and spatial consistency of parameter adjustments.

[0071] The dust removal linkage module's air curtain generator consists of an annular array of nozzles with guide grooves on its inner wall to create a laminar air curtain. The axial drive mechanism is driven by a linear motor, adjusting the nozzle array's extension according to the platform speed, ensuring that the center of the air curtain is always 5-10mm ahead of the laser focus. It should be noted that the air curtain pressure is adjusted by a combination of a proportional valve and a Venturi tube. As the platform speed increases, the cross-sectional area of the Venturi tube's throat automatically decreases to enhance the negative pressure effect, thereby forming a stable dust isolation zone in front of the processing area.

[0072] The benefits of this embodiment include significantly improving the spatiotemporal synchronization accuracy of dynamic on-the-fly processing through the combination of a high-precision encoder and a prediction algorithm. The cross-coupling compensation mechanism of the energy gradient controller effectively balances the conflicting demands of transmittance and linewidth uniformity, while the field of view expansion and error correction of the small BOX splicing module provide technical support for large-scale continuous processing. Furthermore, the speed-following design of the air curtain generator reduces the risk of dust redeposition by pre-setting isolation zones, enhancing process stability.

[0073] Although the present invention has been described above in detail with reference to its preferred embodiments, it is to be understood that the present invention is not limited to the embodiments described above. Instead, various modifications and changes may be made by those skilled in the art without departing from the spirit of the invention, and these modifications and changes should fall within the scope defined by the appended claims and their equivalents.

Claims

1. A method for controlling laser light transmission scribing of solar thin film cells, characterized in that: The control method includes: Through the dynamic flight continuous processing mode, the large-scale substrate is divided into multiple galvanometer fields of view, and a continuous serpentine processing path is generated. The boundaries of adjacent galvanometer fields of view are formed into a splicing area, the laser processing platform is driven to move continuously at a constant speed, and the galvanometer scanning system is synchronously triggered to start scribing in advance in the splicing area; Small BOX format segmentation and splicing control is adopted to divide the entire processing area into multiple independent small BOX units. The range of the galvanometer field of view is dynamically expanded based on the real-time platform displacement. Through the closed-loop matching of encoder feedback and galvanometer scanning speed, the line width error is controlled within ±3μm. During the laser emission phase, the first pulse is suppressed, and ultrafast laser modulation technology is used to nonlinearly attenuate the peak power of the first N pulses, eliminating mura defects caused by abnormal first pulse energy. A transmittance feedback compensation mechanism is established based on the real-time coordinates of the flight path. The transmittance uniformity of the stitching area is synchronously scanned by a line array camera. When the transmittance difference between adjacent stitching areas exceeds a set threshold, the laser parameters and the marking interval are dynamically adjusted. The linked dust removal system follows the motion trajectory of the laser processing head, and the etching position is tracked in real time through an independent axially driven dust extraction device. A high-pressure gas injection module is integrated inside the dust extraction port to form a directional airflow barrier to suppress dust diffusion.

2. The solar thin film cell laser light transmission scribing control method according to claim 1, characterized in that: The steps of the dynamic flight continuous processing mode include: A dynamic expansion model of the galvanometer field of view is established. The expansion amount of the galvanometer field of view is calculated according to the platform movement speed and the maximum scanning angle of the galvanometer to generate an overlapping scanning area covering the adjacent field of view boundaries. The calculation formula for the expansion amount of the galvanometer field of view is: , where t0 is the initial time, t is the continuous processing time, v(t') is the instantaneous speed of the platform at time t', θ max is the maximum scanning angle of the galvanometer; Using a synchronous trigger algorithm, when the laser focus coordinates enter the critical area within 5% of the current field of view boundary, the scanning galvanometer deflection control of the next field of view is started; Energy gradient control is implemented in the overlapping scanning area, and the line width change rate of the overlapping scanning area satisfies the following conditions by coupling adjustment of the laser pulse frequency f and the duty cycle D: , where W is the line width and x represents the coordinate position along the line path; The synchronization triggering algorithm includes: The platform displacement is captured in real time by a high-precision encoder, and the vector coordinate sequence of the serpentine path is preloaded in the galvanometer scanning system; When the platform displacement reaches the field of view switching threshold, the galvanometer deflection controller is triggered to load the scanning parameters of the next field of view in advance. The loading time difference is less than 20% of the system response delay time, and the quadratic interpolation algorithm of the galvanometer scanning angle is used to eliminate the mechanical gap error during the loading process.

3. The solar thin film cell laser light transmission scribing control method according to claim 1, characterized in that: The steps of controlling the small BOX format segmentation and splicing include: A dynamic error compensation model is established based on the small BOX unit size. The influence of the platform motion acceleration a(t) on the splicing accuracy is eliminated by the galvanometer compensation amount. The calculation formula of the galvanometer compensation amount is: , where k is the galvanometer scanning resolution correction coefficient; A dual-weight fusion function is established at the boundaries of adjacent small boxes to perform smooth transition processing on the line coordinates: , where a represents the weight coefficient, x1 and x2 represent the original coordinates of adjacent small boxes, and the weight coefficients satisfy: , where d0 represents the length of the fusion interval.

4. The solar thin film cell laser light transmission scribing control method according to claim 1, characterized in that: The step of first pulse emphasis suppression comprises: Establish the first pulse energy decay function: , where E n is the energy of the nth pulse, E0 is the initial pulse energy, γ is the attenuation coefficient, n is the pulse sequence number, and τ is the time constant; Apply nonlinear power modulation to the first five pulses to satisfy: , n∈[1,5], where P n is the power of the nth pulse, P max is the maximum power.

5. The solar thin film cell laser light transmission scribing control method according to claim 1, characterized in that: The transmittance feedback compensation mechanism includes: A transmittance balancing control algorithm was constructed, using a proportional-integral-differential control strategy to dynamically compensate for real-time transmittance deviations, where the integral term weight is inversely proportional to the platform's motion speed. The marking interval is nonlinearly adjusted according to the compensation amount. The adjustment function is a hyperbolic tangent nonlinear mapping relationship to ensure that the transmittance fluctuation amplitude does not exceed 15% of the initial value. When the platform speed increases to the preset threshold, the marking interval is automatically reduced to maintain transmittance uniformity.

6. The solar thin film cell laser light transmission scribing control method according to claim 1, characterized in that: The control logic of the dust removal following system includes: A dust extraction device position synchronization model was established, and the dust extraction port offset was dynamically adjusted according to the platform movement speed and the dust diffusion angle. The dust diffusion angle was calibrated in real time through particle trajectory analysis using a high-speed camera. A correlation control strategy between the high-pressure gas injection pressure and the platform speed is designed. The injection pressure increases linearly with the square of the speed, and a local negative pressure zone is formed at the front end of the dust extraction port to enhance the dust capture efficiency.

7. The solar thin film cell laser light transmission scribing control method according to claim 3, characterized in that: The dynamic error compensation model includes: The galvanometer scanning angular velocity is pre-compensated according to the platform motion acceleration. An acceleration lag correction factor is introduced when calculating the compensation amount. The value of the correction factor is positively correlated with the torque response time of the platform drive motor. The compensated marking position is closed-loop verified by a laser interferometer.

8. The solar thin film cell laser light transmission scribing control method according to claim 4, characterized in that: The energy decay function includes: A pulse shaping module is set in the laser modulation circuit to apply nonlinear slew rate control to the rising edge of the first pulse. The slew rate adjustment amplitude has a piecewise linear relationship with the preset mura defect elimination level. The pulse energy is monitored in real time by a photodetector and fed back to the dynamic adjustment loop of the attenuation coefficient γ.

9. The solar thin film cell laser light transmission scribing control method according to claim 5, characterized in that: The step of adjusting the line spacing includes: A database of inversely proportional correlation between marking interval and platform speed is established. When the speed reaches a preset critical value, the interval compression mode is activated. The compression ratio is dynamically optimized based on the real-time transmittance sampling data. During the optimization process, a genetic algorithm is used to perform multi-objective optimization of the compression coefficient. The objective function includes the weighted sum of the transmittance uniformity score and the processing efficiency weight.

10. A solar thin film cell laser light transmission scribing control system, characterized in that: The control system includes: A loading and unloading control module is used to fix the glass substrate through a lifting conveyor and a vacuum adsorption mechanism; Dynamic on-the-fly machining module, including a galvanometer scanning subsystem and a motion platform control unit, for achieving continuous serpentine path machining; Small BOX splicing module, used to execute dynamic field of view expansion and error compensation algorithm; Pulse suppression module, integrating pulse shaping circuit and energy monitoring feedback unit; Transmittance closed-loop control module, including a line array camera, an optical sensor, and a real-time parameter adjustment unit; Dust removal linkage module, with an axial drive mechanism and an air curtain generating device with adjustable injection pressure; A central controller, configured to coordinate the modules to execute the operation process of the solar thin-film cell laser light transmission scribing control method according to any one of claims 1 to 9; The dynamic on-the-fly processing module further comprises: High-precision encoder, used to collect platform displacement in real time and convert it into galvanometer deflection control signal; Field of view stitching logic unit with built-in advance triggering algorithm to eliminate mechanical delay. The algorithm predicts the future displacement based on the platform acceleration and loads the galvanometer parameters 3-5 control cycles in advance. The energy gradient controller realizes line width uniformity control by adjusting the pulse frequency and duty cycle of the laser, and the control signal and the output of the transmittance feedback module form a cross-coupling compensation loop.

Images