Intelligent monitoring type BC component integrated preparation method

Abstract

The application provides a kind of intelligent monitoring type BC component integrated preparation method, including the preparation process of positive side napping of BC battery piece, back electrode forming process and passivation film deposition process;Integrated temperature sensor, MCU, voltage sensor and wireless transmission module form a miniature monitoring unit;In the second gap of the back of BC battery piece, the miniature monitoring unit is laid out;Modified epoxy resin potting adhesive is used to potting treatment for miniature monitoring unit;Using the packaging structure of "front glass + packaging adhesive film + BC battery piece + back back plate";The monitoring data is uploaded to the edge computing node through the wireless communication module, the IV curve characteristics of BC component are extracted using the improved convolutional neural network, and the extracted characteristics are intelligently analyzed combined with gradient boosting tree algorithm.The application utilizes the sufficient process space on the back of BC component to realize the integration of miniature monitoring module and component preparation, while ensuring the original photoelectric performance and long-term reliability of the component.

Description

Technical Field

[0001] This invention relates to the field of photovoltaic power generation technology, specifically to an integrated manufacturing method for intelligent monitoring BC modules, which can be widely applied to large-scale ground power stations, distributed rooftop power stations, and other scenarios to achieve full life-cycle health management of photovoltaic modules. Background Technology

[0002] As the core component of a photovoltaic power generation system, the long-term stable operation of photovoltaic modules directly determines the return on investment of the power plant. Existing photovoltaic modules generally lack real-time operating condition monitoring and fault early warning functions. When faults such as microcracks, electrode corrosion, hot spots, and PID effects occur in the modules, they cannot be detected in time, resulting in a 10%-20% decrease in the power generation of a single faulty module. Moreover, the spread of the fault may cause the performance degradation of the entire string or even the entire power station.

[0003] The existing technology has the following drawbacks: (1) Outdated monitoring methods: Traditional photovoltaic power plants rely on manual inspection or drone infrared detection, which has problems such as long detection cycle (the fault location time of distributed roof components usually exceeds 4 hours), high missed detection rate, and delayed fault location.

[0004] (2) Poor compatibility of monitoring modules: Most existing intelligent photovoltaic modules are external, which have defects such as complicated installation, poor compatibility and insufficient weather resistance. Furthermore, they are not designed to adapt to the structural characteristics of BC modules, and cannot give full play to the spatial advantages of BC modules. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an integrated fabrication method for intelligent monitoring BC modules. This method utilizes the ample process space on the back of the BC module to achieve integrated fabrication of the micro-monitoring module and the module itself, solving the problems of lack of monitoring, delayed fault response, and high maintenance costs in traditional modules, while ensuring the original photoelectric performance and long-term reliability of the module.

[0006] The present invention achieves the above objectives through the following technical solutions: A method for integrating and fabricating an intelligent monitoring BC component, comprising: The BC solar cell is subjected to a front surface texturing process, a back electrode formation process, and a passivation film deposition process to ensure that a second gap of a preset width is formed between adjacent back bus regions. The temperature sensor, MCU, voltage sensor and wireless transmission module are integrated into one miniature monitoring unit; the MCU is configured to trigger the temperature sensor at a timed interval to collect the temperature signal of the component backplane at a preset period, and immediately start the continuous acquisition mode when the voltage changes suddenly. In the second gap on the back of the BC solar cell, a micro-monitoring unit is fixed at the center of the gap at a density of one monitoring unit for every 4-6 solar cells, and connected to the solar cell busbar through a flexible wire. Modified epoxy resin potting compound was used to pot the miniature monitoring unit and its connection with the wires. The BC module is formed by adopting a packaging structure of "front glass + encapsulating film + BC cell + back panel", wherein the encapsulating film is EPE co-extruded film. A wireless signal transmission window is reserved at the edge of the BC component, and a signal enhancement antenna is installed. The monitoring data is uploaded to the edge computing node through the wireless communication module. An improved convolutional neural network is used to extract the IV curve features of the BC component, and the extracted features are intelligently analyzed in combination with the gradient boosting tree algorithm to achieve accurate diagnosis and early warning of component failure.

[0007] According to the present invention, a method for integrating and fabricating a smart monitoring type BC module includes performing a front surface texturing process, a back electrode formation process, and a passivation film deposition process on the BC solar cell, comprising the following steps: Front surface texturing preparation: A texturing structure is formed on the front surface of the BC solar cell using chemical etching or laser etching processes; Back electrode formation: An electrode pattern, including main grid lines and sub-grid lines, is formed on the back surface of the BC cell; when forming the electrode pattern, the width and spacing of the electrodes are controlled to reserve a second gap with a width of 800-1000μm between adjacent busbars. Passivation film deposition: A passivation film is deposited on the back surface and electrode pattern of the BC solar cell to reduce surface recombination loss and improve the photoelectric conversion efficiency of the cell; during the deposition of the passivation film, the deposition parameters are controlled to ensure that the passivation film covers the area uniformly and does not block the second gap. Second gap width control: During the formation of the back electrode and the deposition of the passivation film, a high-precision microscope or laser rangefinder is used to monitor and adjust the second gap width between adjacent busbars in real time to ensure that it is always kept within the range of 800-1000μm; if the gap width is detected to deviate from the preset range, the process parameters are immediately adjusted to correct it.

[0008] According to the intelligent monitoring type BC component integrated fabrication method provided by the present invention, a pretreatment step is further performed in the micro monitoring unit: The MCU conditions the weak signal output from the temperature sensor and amplifies the signal using an adjustable amplification factor G; the amplified output voltage U amp Through formula U amp =G×(U in -U ef ) It is calculated that U inU is the original output voltage of the temperature sensor. ef Set the reference voltage to 0.5V; A second-order RC low-pass filter is used to further process the amplified signal. The cutoff frequency fc of the filter is designed to be... ; The filtered digital signal is calibrated, and the digital value is converted into the actual temperature value by using the linear fitting formula T=a×D+b, where a and b are calibration coefficients and D is the digital signal value; at the same time, moving average filtering and outlier removal strategies are implemented.

[0009] According to the intelligent monitoring type BC component integrated fabrication method provided by the present invention, a temperature sensor, an MCU, a voltage sensor and a wireless transmission module are integrated on the same micro-circuit carrier to form a micro-monitoring unit with independent signal processing and transmission functions; wherein, each module is electrically interconnected through a low-power printed circuit board; The MCU is configured to trigger the temperature sensor at a preset period T to collect the temperature signal of the BC component backplane and temporarily store the collected temperature data in the MCU's built-in memory. The voltage sensor monitors the output voltage of the BC component in real time, and immediately sends an interrupt signal to the MCU when a voltage surge event is detected. After receiving an interrupt signal, the MCU automatically switches to continuous acquisition mode, continuously acquiring temperature sensor data at the minimum sampling interval Δt until the voltage sensor detects that the voltage has returned to a stable state, and then resumes the timed acquisition mode. The MCU stores the collected temperature data in a circular buffer after marking it with a timestamp. When the wireless communication module is idle, it encapsulates the buffered data into standard data frames according to the first-in-first-out principle and sends them to the edge computing node through the wireless communication module.

[0010] According to the intelligent monitoring BC module integrated fabrication method provided by the present invention, based on the arrangement of the cells in the BC module, a micro monitoring unit is planned and set in the second gap area on the back of every 4-6 adjacent BC cells; wherein, the second gap is defined as the non-conductive area between the busbar areas on the back of adjacent cells, and its width ranges from 800-1000μm. A high-precision visual positioning system or laser positioning device is used to mark the geometric center point in the selected second gap area as the fixed position of the micro monitoring unit; At the center position after positioning, the miniature monitoring unit is fixed by any of the following methods: conductive adhesive bonding, welding, or mechanical snap-fit. Tinned copper core flexible wires are used to electrically connect the input / output pins of the micro monitoring unit to the battery cell busbar; the connection methods include at least manual soldering, laser soldering, and crimping.

[0011] According to the present invention, a method for preparing an integrated intelligent monitoring BC component is provided, in which modified epoxy resin potting compound is used as the encapsulation material, the prepared potting compound is uniformly poured onto the surface of the micro monitoring unit and the wire connection part, the potting layer thickness is controlled to be 5-10mm, and the thickness is monitored in real time by a laser displacement sensor or a micrometer. The operation shall be carried out in a potting chamber with a cleanliness level of not less than ISO Class 5, with the ambient temperature controlled at 25-30℃ and the vacuum degree ≤5×10⁻⁶. -3 Pa; Pre-degass the potting compound using a vacuum degassing device for ≥15 minutes; In a vacuum environment, a segmented injection method is used. First, the injection is made to 50% of the designed thickness, then it is allowed to stand for 5 minutes before the injection is continued to the final thickness. During the injection process, an ultrasonic oscillator is used to help remove air bubbles. After potting, the components are transferred to a constant temperature curing oven and cured at a preset temperature. During the curing process, the components are kept horizontal to avoid deformation of the potting compound due to gravity. After curing, the surface smoothness of the potting compound is tested using a surface roughness meter, and the arithmetic mean roughness Ra is required to be ≤1μm. For components with excessive surface smoothness, local repair is carried out using precision grinding equipment.

[0012] According to the intelligent monitoring BC component integrated fabrication method provided by the present invention, the following layered packaging structure and lamination process are adopted in the packaging structure and process of the intelligent monitoring BC component, which are arranged from top to bottom as follows: Front cover: Made of ultra-clear patterned tempered glass; Encapsulation film: EPE co-extruded film is used, and its structure is an EVA-POE-EVA sandwich composite layer; Functional layer: BC cell array with integrated micro-monitoring units. The micro-monitoring units are connected to the cell busbars via flexible wires, and the wire connection points are encapsulated with modified epoxy resin. Back panel: TPT composite back panel with double-sided fluorination treatment structure; Encapsulation is performed using a vacuum laminator, and specific process controls include: Temperature control: Set the lamination temperature to 145-150℃, with a three-stage temperature control during the heating phase; Pressure control: Two-stage pressurization: Initial vacuum pressure ≤ 5 × 10 -2 Pa, holding time; During the positive pressure phase, the pressure is 0.3-0.5 MPa, and the pressure increases linearly through a servo hydraulic system. Time control: Total lamination time 15-20 minutes, of which: Vacuuming time ≥ 8 minutes; The pressure should be maintained for 8-12 minutes. Forced air cooling is used during the cooling phase, and the lid can only be opened when the temperature drops below 60℃.

[0013] According to the intelligent monitoring BC component integration preparation method provided by the present invention, in the wireless data transmission method of the intelligent monitoring BC component, Bluetooth Low Energy 5.0 or above is adopted, configured as a non-connected broadcast mode, and the broadcast data packet type is selected as ADV_NONCONN_IND; The monitoring module sends data frames at a preset period. The data frame structure includes at least: preamble, device ID, temperature data, voltage data, status flag, and CRC checksum. When there is a sudden voltage change or the temperature exceeds the limit, it automatically switches to sending emergency data frames at a cycle of 1±0.1s, sets the status flag to 0xFF, and continues to send until the fault is cleared. Enable the adaptive frequency hopping function of the BLE protocol, configure the number of frequency hopping channels to 40, covering the 2.402-2.480GHz frequency band; The frequency hopping interval is set to 1MHz, the frequency hopping map is dynamically updated every 60±5s, and the update is based on the channel quality assessment value. When three consecutive data transmission failures are detected, the channel blacklist mechanism is automatically triggered, and the interfering channel is added to the disabled list. The edge computing nodes are equipped with multi-channel Bluetooth receivers and use time-division multiplexing technology to allocate reception time slots.

[0014] According to the intelligent monitoring type BC component integrated fabrication method provided by the present invention, the IV curve features of the BC component are extracted by an improved convolutional neural network, which is expressed as the following formula;

[0015] Where I is the component operating current, V is the operating voltage, and I sc I0 is the short-circuit current, I0 is the reverse saturation current, q is the electron charge, n is the ideality factor, k is the Boltzmann constant, T is the component temperature, and R0 is the component temperature. s For series resistance, R sh These are parallel resistors; Voltage data and current sampling data collected by the micro-monitoring unit are used to fit and generate the IV curve features of the photovoltaic module. After normalization, a standardized IV curve data matrix is ​​obtained, which is used as the input of the gradient boosting tree algorithm.

[0016] According to the intelligent monitoring-type BC module integrated fabrication method provided by the present invention, a convolutional neural network is used to perform deep feature mining on the IV curve features of the photovoltaic module, and outputs a 64-dimensional feature vector, including the following key parameter derived features: Electrical parameter change rate: Voc deviation rate, Isc attenuation rate, FF decrease rate; Impedance characteristic changes: R_s growth rate, Rsh attenuation rate; Curve morphology characteristics: fill factor curvature, slope at maximum power point; Statistical analysis was performed on temperature data from the past five collection cycles to generate 3D environmental characteristics: The 64-dimensional CNN features are concatenated with the 3-dimensional temperature features to form a 67-dimensional hybrid feature vector, which is used as the input data for the gradient boosting tree model. Constructing a hierarchical fault diagnosis model: Base model selection: The GBDT model was constructed using the XGBoost algorithm; Establish a tiered decision-making mechanism: Phase 1: Initial classification of fault types based on CNN features, outputting 5 candidate fault types; Second stage: Adjust the confidence level of the initial judgment results based on temperature characteristics; The model's final output includes two parts of the diagnostic results: Fault type: 5-dimensional vector using One-Hot encoding; Confidence level: A continuous value between 0 and 1, accurate to 3 decimal places.

[0017] Therefore, compared with the prior art, the intelligent monitoring BC component integrated fabrication method proposed in this invention has the following beneficial effects: 1. Traditional photovoltaic modules lack real-time monitoring capabilities, leading to delayed fault detection and resulting in power generation losses. This invention integrates a micro-monitoring module embedded in the back of the BC module to achieve real-time acquisition and dynamic analysis of key parameters such as temperature and voltage. Combining an improved convolutional neural network (CNN) and gradient boosting tree (GBDT) algorithm, it can accurately identify 14 typical faults, including microcracks, hot spots, and electrode corrosion. This technology reduces fault location time from hours in traditional methods to seconds, preventing fault propagation and subsequent performance degradation of the entire string, thus significantly reducing power generation losses.

[0018] 2. Existing external monitoring modules suffer from cumbersome installation and poor compatibility, and may obstruct the front grid lines of the module or occupy space on the back, affecting photoelectric efficiency. This invention utilizes an 800-1000μm wide gap between the busbar and collector areas on the back of the BC module to achieve integrated monitoring module and module through an embedded design. This design does not require alteration to the front structure of the module, avoids grid line obstruction, and ensures electrical connection stability and encapsulation integrity between the monitoring module and the solar cells through low-damage laser micro-welding and modified epoxy resin potting processes.

[0019] 3. Photovoltaic modules often face extreme environments such as high temperature, high humidity, and low temperature, leading to monitoring module failure or performance degradation. This invention employs a weather-resistant enhancement process. The encapsulating film uses EPE co-extruded film (EVA-POE-EVA sandwich structure), combining the high weather resistance of POE with the process compatibility of EVA. The potting compound uses a highly resistant to humidity and heat formula, and through vacuum potting and 80℃ curing treatment, air bubbles are eliminated and the structural density is improved, enabling the module to adapt to extreme temperatures of -40℃ to 85℃ and high humidity environments of 85%RH. After 1000 hours of humidity and heat aging, the monitoring function still maintains stability, significantly improving the long-term operational reliability of the module under harsh climatic conditions.

[0020] 4. Existing smart module manufacturing requires additional core equipment, leading to a significant increase in production costs. The process of this invention is fully compatible with existing BC module production lines, requiring only the addition of a monitoring module integration and debugging station before lamination, without the need for modification or the addition of key equipment. Through standardized operating procedures and modular design, it can reduce the incremental cost per module and support large-scale industrial production, providing a low-cost, high-efficiency solution for the intelligent upgrading of the photovoltaic industry.

[0021] 5. Traditional photovoltaic power plants rely on manual inspections or drone checks, resulting in low operation and maintenance efficiency and high costs. This invention transforms the operation and maintenance model from reactive maintenance to proactive prevention through real-time monitoring and AI-powered fault early warning. The system can identify potential faults in advance and trigger warnings, reducing unplanned downtime and lowering operation and maintenance costs. Simultaneously, the modular design supports rapid replacement of fault monitoring units, further shortening the maintenance cycle.

[0022] 6. Existing fault diagnosis methods rely solely on electrical parameters (such as IV curves), making them susceptible to environmental interference and prone to misdiagnosis. This invention constructs a multimodal diagnostic model by fusing IV curve features (64 dimensions) with temperature time-series data (3 dimensions). For example, hot spot faults and normal high-temperature fluctuations exhibit similar IV curve morphologies, but by correcting for temperature data (the local temperature of a hot spot fault is 8-20°C higher than the ambient temperature, while high-temperature fluctuations are only 3-5°C higher), the two types of operating conditions can be accurately distinguished, avoiding false alarms and significantly improving diagnostic robustness in complex scenarios.

[0023] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. Attached Figure Description

[0024] Figure 1 This is a flowchart of an embodiment of the intelligent monitoring type BC component integrated preparation method of the present invention.

[0025] Figure 2 This is a schematic diagram of the packaging structure of an embodiment of the intelligent monitoring type BC component integrated fabrication method of the present invention.

[0026] Figure 3 This is a flowchart illustrating the front surface texturing process, back electrode formation process, and passivation film deposition process of a BC solar cell in an embodiment of the intelligent monitoring type BC module integrated fabrication method of the present invention.

[0027] Figure 4 This is a flowchart of the preprocessing steps of the micro-monitoring unit in an embodiment of the intelligent monitoring type BC component integrated fabrication method of the present invention.

[0028] Figure 5 This is a physical schematic diagram of the BC component in an embodiment of the intelligent monitoring type BC component integrated preparation method of the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0030] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0031] See Figures 1 to 5 This embodiment provides a method for integrating and fabricating an intelligent monitoring BC component, including: Step S1: Perform front surface texturing process, back electrode formation process and passivation film deposition process on BC solar cells to ensure that a second gap of a preset width is formed between adjacent back busbars. Step S2: Integrate the temperature sensor, MCU, voltage sensor and wireless transmission module into one unit to form a miniature monitoring unit; wherein, the MCU is configured to trigger the temperature sensor at a timed interval to collect the component backplane temperature signal at a preset period, and immediately start the continuous acquisition mode when the voltage changes suddenly. Step S3: In the second gap on the back of the BC solar cell, a micro-monitoring unit is fixed at the center of the gap at a density of one monitoring unit for every 4-6 solar cells, and connected to the solar cell busbar through a flexible wire. Step S4: The micro monitoring unit and its connection with the wire are encapsulated using modified epoxy resin potting compound. Step S5: The BC module is formed by adopting the encapsulation structure of "front glass + encapsulation film + BC cell + back panel", wherein the encapsulation film is EPE co-extruded film. Step S6: Reserve a wireless signal transmission window at the edge of the BC component and install a signal enhancement antenna. Upload the monitoring data to the edge computing node through the wireless communication module. Use an improved convolutional neural network to extract the IV curve features of the BC component and combine the gradient boosting tree algorithm to perform intelligent analysis on the extracted features, so as to achieve accurate diagnosis and early warning of component failure.

[0032] In step S1 above, as Figure 3 As shown, the process of fabricating the front surface of the BC solar cell, forming the back electrode, and depositing the passivation film includes the following steps: Front surface texture preparation: A textured structure is formed on the front surface of the BC solar cell using chemical etching or laser etching processes to increase the light absorption area and reduce reflection. The average roughness of the textured structure is controlled within a preset range to optimize optical performance. Back electrode formation: On the back surface of the BC cell, electrode patterns, including main grid lines and sub-grid lines, are formed by methods such as screen printing, vapor deposition or electroplating to ensure good ohmic contact between the electrodes and the silicon substrate of the cell; When forming the electrode pattern, the width and spacing of the electrodes are precisely controlled to reserve a second gap with a width of 800-1000μm between adjacent busbars. Passivation film deposition: A passivation film is deposited on the back surface and electrode pattern of the BC solar cell using methods such as chemical vapor deposition (CVD), physical vapor deposition (PVD), or atomic layer deposition (ALD). The passivation film material is selected from silicon nitride, aluminum oxide, or a combination thereof, and is used to reduce surface recombination loss and improve the photoelectric conversion efficiency of the solar cell. During the deposition of the passivation film, the deposition parameters are strictly controlled to ensure that the passivation film is uniformly covered and does not block the second gap. Second gap width control: During the formation of the back electrode and the deposition of the passivation film, the second gap width between adjacent busbars is monitored and adjusted in real time through process control and detection methods, such as using a high-precision microscope or laser rangefinder, to ensure that it is always kept within the range of 800-1000μm; if the gap width is detected to deviate from the preset range, the process parameters (such as printing speed, deposition rate, etc.) are immediately adjusted for correction.

[0033] In this embodiment, as Figure 4 As shown, a preprocessing step is also performed in the micro-monitoring unit: The MCU conditions the weak signal (voltage range 0-1V) output from the temperature sensor and amplifies it using an adjustable amplification factor G (G's value ranges from [100, 1000]) to improve the signal-to-noise ratio; the amplified output voltage U amp Through formula U amp =G×(U in -U ef ) It is calculated that U in U is the original output voltage of the temperature sensor. ef Set the reference voltage to 0.5V; A second-order RC low-pass filter is used to further process the amplified signal. The cutoff frequency fc of the filter is designed to be... For a frequency of 10Hz, a combination of resistor R=10kΩ and capacitor C=1.6μF can be selected to effectively filter out electromagnetic interference in the 1kHz-1MHz frequency band and ensure that the signal-to-noise ratio (SNR) of the processed signal is not less than 60dB. The filtered digital signal is calibrated and converted into an actual temperature value using the linear fitting formula T=a×D+b, where a and b are calibration coefficients and D is the digital signal value. At the same time, a moving average filter (using 5 data windows) and an outlier removal strategy (discarding data points when the temperature value exceeds the range of -50~100℃) are implemented to improve the accuracy and stability of temperature measurement. The models and parameters of the components used are as follows: Temperature sensor: Thin-film adhesive type, with a measurement range of -50~100℃, an accuracy of ±0.5℃, and a power consumption of no more than 0.15W; Voltage sensor: Measurement accuracy is ±0.1V, and it can be adapted to the open circuit voltage range of the BC component; Wireless transmission module: It adopts the CC2340 Bluetooth module, supports the Bluetooth 5.0 protocol, the module size does not exceed 5×5×2mm, uses DC5V power supply, and achieves reliable electrical connection with the back electrode of the BC component through a miniature flexible wire.

[0034] In step S2 above, the temperature sensor, MCU, voltage sensor and wireless transmission module are integrated on the same micro circuit carrier to form a micro monitoring unit with independent signal processing and transmission functions; wherein, each module is electrically interconnected through a low power printed circuit board (PCB) and miniaturized packaging technology is used to ensure that the overall size is adapted to the installation space requirements of the back gap of the BC component. The MCU is configured to trigger the temperature sensor at a preset period T to collect the temperature signal of the BC component backplane and temporarily store the collected temperature data in the MCU's built-in memory; where T can be set to any value in the range of 1-600 seconds, with a default value of 60 seconds.

[0035] The voltage sensor monitors the output voltage of the BC component in real time. When a voltage change event is detected, an interrupt signal is immediately sent to the MCU. The voltage change event is defined as the voltage change rate exceeding ±5V / second or the absolute value change exceeding a preset threshold ΔV, where ΔV∈[1V, 10V].

[0036] After receiving a voltage surge interrupt signal, the MCU automatically switches to continuous acquisition mode and continuously acquires temperature sensor data at the minimum sampling interval Δt (Δt≤1 second) until the voltage sensor detects that the voltage has returned to a stable state, and then resumes the timed acquisition mode. The stable state of voltage is defined as the voltage change rate being less than ±0.5V / second for three consecutive sampling periods.

[0037] The MCU stores the collected temperature data in a circular buffer after marking it with a timestamp. When the wireless communication module is idle, it encapsulates the buffered data into a standard data frame (containing temperature value, collection timestamp, and component identifier) ​​according to the first-in-first-out (FIFO) principle and sends it to the edge computing node through the wireless communication module.

[0038] In step S3 above, according to the arrangement of the cells in the BC module, a micro monitoring unit is planned to be installed in the second gap area on the back of every 4-6 adjacent BC cells; wherein, the second gap is defined as the non-conductive area between the busbar areas on the back of adjacent cells, and its width ranges from 800-1000μm. A high-precision visual positioning system or laser positioning device is used to mark the geometric center point in the selected second gap area as the fixed position of the micro monitoring unit; the positioning accuracy must be within ±50μm to ensure a safe distance between the monitoring unit and the edge of the battery cell; Once the positioning is complete, fix the miniature monitoring unit at the center location using any of the following methods: Conductive adhesive bonding: Anisotropic conductive adhesive or silver paste conductive adhesive is used to form mechanical fixation and electrical conductivity between the bottom of the monitoring unit and the back of the battery cell; Welding and fixing: The pins of the monitoring unit are welded to the pre-plated pads on the back of the battery cell using low-temperature alloy solder; Mechanical snap-fit: A miniature snap-fit ​​structure is designed to snap the monitoring unit onto the edges of the battery cells on both sides of the second gap through elastic deformation; A tin-plated copper core flexible wire with a diameter ≤0.2mm is used to electrically connect the input / output pins of the micro-monitoring unit to the battery cell busbar; the connection methods include: Manual soldering: Soldering the wires to the busbars using a miniature soldering iron at a temperature of 350±10℃; Laser welding: Non-contact welding of wires and busbars is achieved using pulsed lasers with a wavelength of 1064nm and a pulse width of ≤1ms; Crimp connection: The wires are pressed tightly onto the surface of the busbar using spring-loaded crimp terminals; After the installation and connection are completed, the reliability of the electrical connections is verified through the following tests: Continuity test: Use a micro-ohmmeter to measure the contact resistance between the monitoring unit and the busbar. The resistance value should be ≤10mΩ. Tensile test: Apply an axial tensile force of 5N to the flexible conductor for 10 seconds and check whether the connection point is loose or detached; Thermal cycling test: Place the component in a temperature cycling environment from -40°C to +85°C for 100 cycles, and check whether the resistance change rate at the connection point exceeds 20%.

[0039] In step S4 above, a modified epoxy resin potting compound is used as the encapsulation material. This potting compound must meet the following performance parameters: Viscosity (25℃): 5000-15000 mPa·s; Pot life after mixing (25℃): ≥30 minutes; Volume resistivity: ≥1×10 14 Ω·cm; Glass transition temperature (Tg): ≥120℃; Linear shrinkage rate: ≤0.3%; The prepared potting compound is evenly poured onto the surface of the micro monitoring unit and wire connection parts, and the thickness of the potting layer is controlled to be 5-10mm. The thickness is monitored in real time by a laser displacement sensor or micrometer. The operation shall be carried out in a potting chamber with a cleanliness level of not less than ISO Class 5, with the ambient temperature controlled at 25-30℃ and the vacuum degree ≤5×10⁻⁶. -3 Pa; Pre-degass the potting compound using a vacuum degassing device for ≥15 minutes; In a vacuum environment, a segmented injection method is used. First, the injection is carried out to 50% of the designed thickness, and after standing for 5 minutes, the injection is continued to the final thickness. During the injection process, an ultrasonic oscillator (frequency 20-40kHz) is used to assist in the removal of air bubbles. After potting, the components are transferred to a constant temperature curing oven and cured at 80±1℃ for 2-3 hours. During the curing process, the components are kept horizontal to avoid deformation of the potting compound due to gravity. After curing, the surface smoothness of the potting compound is tested using a surface roughness meter. The arithmetic mean roughness Ra is required to be ≤1μm. For components with excessive surface smoothness, local repair is carried out using precision grinding equipment (diamond grinding paste with a particle size ≤W20). The following performance tests were performed on the potted components: Insulation performance: The insulation resistance between the potting compound and the conductive parts is measured at 500V DC voltage and is required to be ≥100MΩ; Weather resistance: After placing the component in a humid heat environment of 85℃ / 85%RH for 72 hours, check whether the potting compound shows cracking, discoloration or peeling. Mechanical strength: Apply a 10N impact force (0.5J impact energy) to the potting compound and check for any damage or detachment.

[0040] In step S5 above, the following stacked packaging structure and matching lamination process are adopted in the packaging structure and process of the intelligent monitoring BC component: like Figure 2 As shown, the settings from top to bottom are as follows: Front cover: Made of ultra-clear patterned tempered glass with a light transmittance of ≥91% and a thickness of 3.2±0.2mm; Encapsulation film: EPE co-extruded film is used, with an EVA-POE-EVA sandwich composite structure, wherein the POE intermediate layer accounts for ≥60% of the thickness, and the water vapor permeability is ≤1×10⁻⁶. -13 g / (cm 2 ·s); Functional layer: BC cell array with integrated micro monitoring units. The monitoring module is connected to the cell busbar via flexible wires, and the wire connection is encapsulated with modified epoxy resin. Back panel: TPT composite back panel (polyvinyl fluoride composite film) is used, with a thickness range of 0.5±0.1mm and a double-sided fluorinated structure; The encapsulation process is carried out using a vacuum laminator, and the specific process control is as follows: Temperature control: Set the lamination temperature to 145-150℃, with a three-stage temperature control during the heating phase: First stage: Keep warm at 60℃ for 3 minutes; Second stage: Keep warm at 120℃ for 5 minutes; Third stage: Maintain the target temperature until lamination is complete; Pressure control: Two-stage pressurization: Initial vacuum pressure ≤ 5 × 10 -2 Pa, holding time; During the positive pressure phase, the pressure is 0.3-0.5 MPa, and the pressure increases linearly through a servo hydraulic system. Time control: Total lamination time 15-20 minutes, of which: Vacuuming time ≥ 8 minutes; The pressure should be maintained for 8-12 minutes. Forced air cooling is used during the cooling phase; the lid can only be opened when the temperature drops below 60℃. The following protective measures should be implemented during the lamination process: Pressure buffer: A silicone buffer pad (Shore hardness 30-40A) is pre-embedded in the EPE film at the corresponding position of the monitoring module, and its thickness is 0.5mm higher than that of the monitoring module; Temperature monitoring: The surface temperature of the monitoring module is monitored in real time by embedded thermocouples. When the temperature exceeds 135℃, the laminator cooling program is automatically triggered. Structural reinforcement: A polyimide (PI) protective frame is installed around the monitoring module, with a temperature resistance rating of ≥260℃ and an adhesion strength to the EPE film of ≥15N / cm.

[0041] In step S6 above, in the wireless data transmission method of the intelligent monitoring BC component, Bluetooth Low Energy (BLE) 5.0 or above is used, configured as non-connected state broadcast mode, and the broadcast data packet type is selected as ADV_NONCONN_IND; The broadcast interval is set to 50±5ms, the transmission power is 0dBm, the effective transmission distance is ≥20m, and the broadcast channel is selected from three standard channels: 37, 38, and 39. The monitoring module sends data frames at a period of 10 ± 0.5 seconds. The data frame structure includes: Preamble (8bit): 0xAA; Device ID (16-bit): A unique identifier; Temperature data (16-bit): resolution 0.1℃, range -40℃ to +125℃; Voltage data (16-bit): resolution 0.01V, range 0~1000V; Status flag (8 bits): General status flag (0x00); CRC checksum (16-bit): CRC-16 / CCITT-FALSE algorithm; When there is a sudden voltage change (rate of change > ±5V / s) or the temperature exceeds the limit (>90℃), it automatically switches to sending emergency data frames at a cycle of 1±0.1s, sets the status flag to 0xFF, and continues to send until the fault is cleared. Enable the adaptive frequency hopping (AFH) function of the BLE protocol, configure the number of frequency hopping channels to 40, covering the 2.402-2.480GHz frequency band; The frequency hopping interval is set to 1MHz, the frequency hopping map is dynamically updated every 60±5s, and the update is based on the channel quality assessment (CQI) value. When three consecutive data transmission failures are detected, the channel blacklist mechanism is automatically triggered, and the interfering channel (RSSI < -70dBm) is added to the disabled list; The edge computing node is equipped with a multi-channel Bluetooth receiver, which supports the simultaneous reception of data from 8 monitoring modules and uses time division multiplexing (TDM) technology to allocate reception time slots; By extracting the IV curve features of the BC component using an improved convolutional neural network, 14 typical faults such as microcracks, hot spots, and electrode corrosion can be identified, as expressed in the following formula;

[0042] Where I is the component operating current (A), V is the operating voltage (V), and I sc I0 is the short-circuit current (A), I0 is the reverse saturation current (A), q is the electron charge, n is the ideality factor (1~2), k is the Boltzmann constant (1.38×10-23 J / K), T is the module temperature (K), and R0 is the short-circuit current (A). s R is the series resistance (Ω). sh The parallel resistance (Ω) is used; voltage data and current sampling data collected by the micro-monitoring unit are fitted to generate the IV curve characteristics of the photovoltaic module, and then normalized (the voltage and current are normalized to the open circuit voltage V). oc Short-circuit current I sc The range is calculated to obtain a standardized IV curve data matrix (dimension: 1×200, i.e., 200 sampling points), which is used as input for the gradient boosting tree algorithm; The system receives a standardized IV curve data matrix (1×200), uses a convolutional neural network (CNN) to extract local features of the photovoltaic module IV curve (such as curve slope changes and inflection point positions), retains key features and reduces data dimensionality, and outputs a 64-dimensional feature vector.

[0043] A convolutional neural network (CNN) is used to perform deep feature mining on the IV curve features of photovoltaic modules, outputting a 64-dimensional feature vector, including the following key parameter-derived features: Electrical parameter change rate: Voc deviation rate (ΔVoc / Voc_ref), Isc attenuation rate (ΔIsc / Isc_ref), FF decrease rate (ΔFF / FF_ref); Impedance characteristic changes: R_s growth rate (ΔR_s / R_s_ref), Rsh attenuation rate (ΔRsh / Rsh_ref); Curve morphology characteristics: fill factor curvature, slope at maximum power point; Statistical analysis was performed on temperature data from the past five collection cycles to generate 3D environmental characteristics: Average temperature (T_avg): the arithmetic mean of temperatures over 5 periods; Temperature fluctuation (T_var): the standard deviation of temperature over 5 periods; Temperature change rate (T_rate): The temperature difference between the current period and the initial period / time interval; The 64-dimensional CNN features are concatenated with the 3-dimensional temperature features to form a 67-dimensional hybrid feature vector, which is used as the input data for the Gradient Boosting Tree (GBDT) model. Constructing a hierarchical fault diagnosis model: Base model selection: The GBDT model is constructed using the XGBoost algorithm, and the model parameters are set as follows: Number of decision trees: 100; Maximum depth of a single tree: ≤8 layers; Learning rate: 0.05; Subsampling ratio: 0.8; Feature sampling ratio: 0.7; Establish a tiered decision-making mechanism: Phase 1 (Preliminary electrical feature assessment): Based on CNN features, the fault types are initially classified and 5 candidate fault types are output (normal, hot spot, microcrack, PID, connection failure). The second stage (temperature feature correction): The confidence level of the initial judgment is adjusted based on temperature features. The specific correction rules are as follows: When T_avg > 85℃ and T_var < 2℃, reduce the confidence level of "hot spot fault" by 20%; When T_rate > 5℃ / min and T_avg < 60℃, increase the confidence level of "connection failure" by 15%; Output format: The model's final output includes two parts of the diagnostic results: Fault type: 5-dimensional vector using One-Hot encoding; Confidence level: A continuous value between 0 and 1, accurate to 3 decimal places.

[0044] As can be seen, the GBDT algorithm provided in this embodiment constructs 100 regression decision trees to perform hierarchical decision-making on the input features: first, the fault type is initially determined by the IV curve features, as shown in Table 1; then, the diagnostic results are corrected by combining temperature data (such as distinguishing between "hot spot fault" and "normal high temperature fluctuation"), as shown in Table 2; finally, the fault diagnosis results and confidence levels are output.

[0045] Table 1: IV Curve Characteristics and Preliminary Fault Type Determination

[0046] Table 2: Temperature Data Correction Diagnostic Results

[0047] In practical applications, under standard test conditions (STC, AM1.5 spectrum, 1000W / m²), 2 Under irradiance and a module temperature of 25°C, the output power, open-circuit voltage, short-circuit current, and fill factor of the module were measured using an IV tester to verify that the module efficiency degradation rate was ≤0.3%. The surface temperature distribution of the component was scanned using an infrared thermal imager to confirm the absence of localized hot spots (temperature deviation ≤3℃); the component was placed in a temperature change cycle test chamber ranging from -40℃ to 85℃ and subjected to 1000 temperature cycles (heating / cooling rate ≥5℃ / min per cycle). After the test, the module's functional stability was monitored to be ≥98%. Run continuously for 1000 hours in a damp heat aging test chamber at 85℃ and 85%RH to check for delamination, bubbles, or discoloration of the encapsulating film (EPE co-extruded film) and potting compound (high damp heat resistant formula) to confirm that there is no structural failure. Simulated microcrack fault: Microcracks with a width of ≤50μm were created on the surface of the battery cell by laser scribing to verify that the monitoring module triggered a fault warning within 10 seconds and the fault location error was ≤1 battery cell. Simulated hot spot fault: A local area of ​​the component is blocked (blocked area ≤ 5%), and the monitoring module is verified to identify the temperature abnormality (temperature deviation ≥ 8℃) and output an early warning signal within 5 seconds; Long-term stability test: continuous operation for 720 hours, data acquisition frequency of monitoring module ≥ 1 time / minute, data transmission packet loss rate ≤ 0.1%, and AI diagnostic model identification accuracy of ≥ 97% for 14 typical faults.

[0048] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0049] The above embodiments are merely preferred embodiments of the present invention and should not be construed as limiting the scope of protection of the present invention. Any non-substantial changes and substitutions made by those skilled in the art based on the present invention shall fall within the scope of protection claimed by the present invention.

Claims

1. A method for integrating and fabricating an intelligent monitoring BC component, characterized in that, include: The BC solar cell is subjected to a front surface texturing process, a back electrode formation process, and a passivation film deposition process to ensure that a second gap of a preset width is formed between adjacent back bus regions. The temperature sensor, MCU, voltage sensor and wireless transmission module are integrated into one miniature monitoring unit; the MCU is configured to trigger the temperature sensor at a timed interval to collect the temperature signal of the component backplane at a preset period, and immediately start the continuous acquisition mode when the voltage changes suddenly. In the second gap on the back of the BC solar cell, a micro-monitoring unit is fixed at the center of the gap at a density of one monitoring unit for every 4-6 solar cells, and connected to the solar cell busbar through a flexible wire. Modified epoxy resin potting compound was used to pot the miniature monitoring unit and its connection with the wires. The BC module is formed by adopting a packaging structure of "front glass + encapsulating film + BC cell + back panel", wherein the encapsulating film is EPE co-extruded film. A wireless signal transmission window is reserved at the edge of the BC component, and a signal enhancement antenna is installed. The monitoring data is uploaded to the edge computing node through the wireless communication module. An improved convolutional neural network is used to extract the IV curve features of the BC component, and the extracted features are intelligently analyzed in combination with the gradient boosting tree algorithm to achieve accurate diagnosis and early warning of component failure.

2. The method according to claim 1, characterized in that, The process of fabricating the front surface of the BC solar cell, forming the back electrode, and depositing the passivation film includes the following steps: Front surface texturing preparation: A texturing structure is formed on the front surface of the BC solar cell using chemical etching or laser etching processes; Back electrode formation: An electrode pattern, including main grid lines and sub-grid lines, is formed on the back surface of the BC cell; when forming the electrode pattern, the width and spacing of the electrodes are controlled to reserve a second gap with a width of 800-1000μm between adjacent busbars. Passivation film deposition: A passivation film is deposited on the back surface and electrode pattern of the BC solar cell to reduce surface recombination loss and improve the photoelectric conversion efficiency of the cell; during the deposition of the passivation film, the deposition parameters are controlled to ensure that the passivation film covers the area uniformly and does not block the second gap. Second gap width control: During the formation of the back electrode and the deposition of the passivation film, a high-precision microscope or laser rangefinder is used to monitor and adjust the second gap width between adjacent busbars in real time to ensure that it is always kept within the range of 800-1000μm; if the gap width is detected to deviate from the preset range, the process parameters are immediately adjusted to correct it.

3. The method according to claim 1, characterized in that, In the micro-monitoring unit, a preprocessing step is also performed: The MCU conditions the weak signal output from the temperature sensor and amplifies the signal using an adjustable amplification factor G; the amplified output voltage U amp Through formula U amp =G×(U in -U ef ) It is calculated that U in U is the original output voltage of the temperature sensor. ef Set the reference voltage to 0.5V; A second-order RC low-pass filter is used to further process the amplified signal. The cutoff frequency fc of the filter is designed to be... ; The filtered digital signal is calibrated, and the digital value is converted into the actual temperature value by using the linear fitting formula T=a×D+b, where a and b are calibration coefficients and D is the digital signal value; at the same time, moving average filtering and outlier removal strategies are implemented.

4. The method according to claim 1, characterized in that: Temperature sensors, MCUs, voltage sensors, and wireless transmission modules are integrated onto the same micro-circuit carrier to form a micro-monitoring unit with independent signal processing and transmission functions; the modules are electrically interconnected through low-power printed circuit boards. The MCU is configured to trigger the temperature sensor at a preset period T to collect the temperature signal of the BC component backplane and temporarily store the collected temperature data in the MCU's built-in memory. The voltage sensor monitors the output voltage of the BC component in real time, and immediately sends an interrupt signal to the MCU when a voltage surge event is detected. After receiving an interrupt signal, the MCU automatically switches to continuous acquisition mode, continuously acquiring temperature sensor data at the minimum sampling interval Δt until the voltage sensor detects that the voltage has returned to a stable state, and then resumes the timed acquisition mode. The MCU stores the collected temperature data in a circular buffer after marking it with a timestamp. When the wireless communication module is idle, it encapsulates the buffered data into standard data frames according to the first-in-first-out principle and sends them to the edge computing node through the wireless communication module.

5. The method according to claim 1, characterized in that: Based on the arrangement of the solar cells in the BC module, a micro monitoring unit is planned to be installed in the second gap area on the back of every 4-6 adjacent BC solar cells; wherein, the second gap is defined as the non-conductive area between the busbar areas on the back of adjacent solar cells, and its width ranges from 800-1000μm. A high-precision visual positioning system or laser positioning device is used to mark the geometric center point in the selected second gap area as the fixed position of the micro monitoring unit; At the center position after positioning, the miniature monitoring unit is fixed by any of the following methods: conductive adhesive bonding, welding, or mechanical snap-fit. Tinned copper core flexible wires are used to electrically connect the input / output pins of the micro monitoring unit to the battery cell busbar; the connection methods include at least manual soldering, laser soldering, and crimping.

6. The method according to claim 2, characterized in that: Modified epoxy resin potting compound is used as the encapsulation material. The prepared potting compound is uniformly poured onto the surface of the micro monitoring unit and the wire connection parts. The thickness of the potting layer is controlled to be 5-10mm. The thickness is monitored in real time by a laser displacement sensor or a micrometer. The operation shall be carried out in a potting chamber with a cleanliness level of not less than ISO Class 5, with the ambient temperature controlled at 25-30℃ and the vacuum degree ≤5×10⁻⁶. -3 Pa; Pre-degass the potting compound using a vacuum degassing device for ≥15 minutes; In a vacuum environment, a segmented injection method is used. First, the injection is made to 50% of the designed thickness, then it is allowed to stand for 5 minutes before the injection is continued to the final thickness. During the injection process, an ultrasonic oscillator is used to help remove air bubbles. The potted components are transferred to a constant temperature curing oven and cured at a preset temperature. Keep the components horizontal during the curing process.

7. The method according to claim 1, characterized in that: In the packaging structure and process of the intelligent monitoring BC component, the following layered packaging structure and lamination process are adopted, arranged from top to bottom as follows: Front cover: Made of ultra-clear patterned tempered glass; Encapsulation film: EPE co-extruded film is used, and its structure is an EVA-POE-EVA sandwich composite layer; Functional layer: BC cell array with integrated micro-monitoring units. The micro-monitoring units are connected to the cell busbars via flexible wires, and the wire connection points are encapsulated with modified epoxy resin. Back panel: TPT composite back panel with double-sided fluorination treatment.

8. The method according to any one of claims 1 to 7, characterized in that: In the wireless data transmission method of the intelligent monitoring BC component, Bluetooth Low Energy 5.0 or above is used, configured as a non-connected broadcast mode, and the broadcast data packet type is selected as ADV_NONCONN_IND; The monitoring module sends data frames at a preset period. The data frame structure includes at least: preamble, device ID, temperature data, voltage data, status flag, and CRC checksum. When there is a sudden voltage change or the temperature exceeds the limit, it automatically switches to sending emergency data frames at a cycle of 1±0.1s, sets the status flag to 0xFF, and continues to send until the fault is cleared. Enable the adaptive frequency hopping function of the BLE protocol, configure the number of frequency hopping channels to 40, covering the 2.402-2.480GHz frequency band; The frequency hopping interval is set to 1MHz, the frequency hopping map is dynamically updated every 60±5s, and the update is based on the channel quality assessment value. When three consecutive data transmission failures are detected, the channel blacklist mechanism is automatically triggered, and the interfering channel is added to the disabled list. The edge computing nodes are equipped with multi-channel Bluetooth receivers and use time-division multiplexing technology to allocate reception time slots.

9. The method according to any one of claims 1 to 7, characterized in that: The IV curve features of the BC component are extracted using an improved convolutional neural network, as expressed by the following formula; Where I is the component operating current, V is the operating voltage, and I sc I0 is the short-circuit current, I0 is the reverse saturation current, q is the electron charge, n is the ideality factor, k is the Boltzmann constant, T is the component temperature, and R0 is the short-circuit current. s For series resistance, R sh These are parallel resistors; Voltage data and current sampling data collected by the micro-monitoring unit are used to fit and generate the IV curve features of the photovoltaic module. After normalization, a standardized IV curve data matrix is ​​obtained, which is used as the input of the gradient boosting tree algorithm.

10. The method according to claim 9, characterized in that: A convolutional neural network is used to perform deep feature mining on the IV curve features of photovoltaic modules, outputting a 64-dimensional feature vector, including the following key parameter-derived features: Electrical parameter change rate: Voc deviation rate, Isc attenuation rate, FF decrease rate; Impedance characteristic changes: R_s growth rate, Rsh attenuation rate; Curve morphology characteristics: fill factor curvature, slope at maximum power point; Statistical analysis was performed on temperature data from the past five collection cycles to generate 3D environmental characteristics: The 64-dimensional CNN features are concatenated with the 3-dimensional temperature features to form a 67-dimensional hybrid feature vector, which is used as the input data for the gradient boosting tree model.

Images