Distributed solar power supply and wireless networking system

Abstract

The invention discloses a distributed solar power supply and wireless networking system, and relates to the technical field of distributed solar power supply networking, and the system comprises a metamaterial composite heat collection module which uses a gradient nanorod array and a phase change material to improve the power generation efficiency; the self-adaptive energy storage module optimizes energy management by means of a honeycomb ceramic container and bidirectional conversion; the dynamic topology networking module is used for constructing a hybrid topology based on a block chain and shortening communication delay; and the intelligent power scheduling module is used for improving the energy utilization rate by using a reinforcement learning algorithm, and further comprises a self-repairing photovoltaic unit and an environment self-adaptive heat dissipation and intelligent fault diagnosis module to realize system self-maintenance and fault early warning. The system is advantaged in that daily generating capacity is improved, fault diagnosis accuracy is improved, battery life is prolonged, self-repairing capability is strong, maintenance cost is low, and the system is suitable for remote area power supply and Internet of Things nodes.

Description

Technical Field

[0001] This invention relates to the field of distributed solar power grid technology, and more particularly to a distributed solar power grid and wireless networking system. Background Technology

[0002] In the field of distributed energy, traditional solar power systems mostly use a combination of monocrystalline silicon solar panels and lead-acid batteries. Limited by material properties and structural design, these systems suffer from low light capture efficiency and inadequate thermal management of energy storage units. Such systems experience unstable power generation when sunlight intensity fluctuates, leading to significant power supply interruptions at night or on cloudy / rainy days. This is particularly problematic in remote areas with insufficient grid coverage, often requiring supplementary diesel generators, resulting in high operating costs and significant environmental pollution. Furthermore, traditional systems lack effective temperature control mechanisms, causing significant battery capacity degradation at high temperatures. Within an ambient temperature range of -20℃ to 60℃, capacity fluctuations can exceed 40%, severely impacting power supply reliability.

[0003] Existing wireless networking technologies have significant drawbacks when applied to distributed solar power systems. Most systems employ fixed topologies, with node communication relying on a single protocol. When the number of nodes exceeds 20, communication congestion easily occurs, and data transmission latency exceeds 100ms, failing to meet real-time monitoring requirements. The application of blockchain technology in energy networking remains largely theoretical. Existing consensus algorithms are energy-intensive and have long consensus cycles, making them difficult to adapt to low-power solar nodes, resulting in delays in uploading device status data to the blockchain and lagging fault warnings. Furthermore, traditional networking systems lack adaptive topology adjustment capabilities; when a node runs out of energy or the communication link quality deteriorates, the network cannot be dynamically reconstructed, posing a single point of failure risk.

[0004] The development of intelligent management technology for solar power systems lags behind, making it difficult to adapt to complex environmental changes. Traditional power dispatching often relies on rule engines, which cannot handle the multi-dimensional coupling relationships between sunlight intensity, load demand, and energy storage status, resulting in energy utilization rates generally below 80%. In multi-node distributed scenarios, each unit operates independently, lacking a global optimization mechanism, often leading to a coexistence of local overload and energy waste. Furthermore, system fault diagnosis depends on manual inspections, with a lack of monitoring of key parameters such as vibration and temperature, resulting in delays in detecting equipment anomalies exceeding 24 hours, causing fault propagation and increased maintenance costs. Existing self-healing technologies have limited application in photovoltaic modules; efficiency degradation caused by surface scratches or coating aging cannot be automatically repaired, resulting in an average annual efficiency loss of over 5%. Summary of the Invention

[0005] The distributed solar power supply and wireless networking system proposed in this invention aims to solve the problems mentioned in the prior art.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: A distributed solar power supply and wireless networking system includes the following modules: Metamaterial composite solar collector module: A gradient refractive index nanopillar array composed of titanium dioxide and silicon is integrated on the surface of a solar panel. The structure is constructed using the principle of multilayer film interference. A phase change material composed of octadecane and expanded graphite is applied to the back of the solar collector. The thermal conductivity is improved by using a graphene network. A thermoelectric generator is embedded at the edge of the module. The temperature difference between the solar collector and the environment is set to >12℃ as the power generation trigger condition. Adaptive energy storage and energy management module: It adopts a honeycomb ceramic thermal storage container, filled with decanoic acid-lauric acid eutectic phase change material, and the inner wall of the container is coated with molybdenum disulfide nano-coating. The energy storage system integrates supercapacitors and lithium batteries, is equipped with a bidirectional DC-DC converter, and designs an energy management algorithm to adjust the charging and discharging strategy according to light intensity and load demand. Dynamic Topology Wireless Networking Module: Constructs a blockchain-based distributed networking protocol. Each node is equipped with both ZigBee and LoRa communication modules. The networking algorithm adjusts the topology by optimizing link quality and energy consumption. The link quality evaluation formula is as follows: , For the link quality between nodes i and j, To receive signal strength, The remaining energy of node i. As initial energy, , These are the weighting coefficients; Intelligent power scheduling module: Deploys light, temperature, and load current sensors; uses an STM32H750 main control chip; runs a deep reinforcement learning algorithm to optimize power allocation; the action space is the output power of each node; the scheduling objective function is: in, For load power, For energy loss, For total energy, These are the weighting coefficients.

[0007] Furthermore, it also includes a self-healing photovoltaic unit module: the surface of the solar panel is sprayed with a dopamine-silver nanoparticle composite coating, which self-heals through a biomimetic mussel adhesion mechanism. The module has a built-in microelectromechanical system micromirror array, and when the battery efficiency in a certain area drops by more than 15%, the micromirrors automatically adjust their angle to focus the light energy.

[0008] Furthermore, it also includes an environmentally adaptive heat dissipation module: a temperature-sensitive and light-sensitive dual-response coating is prepared on the surface of the energy storage container, with a bottom layer of polyisopropylacrylamide hydrogel and a top layer of titanium dioxide nanotubes. When the temperature is >32℃ or the light intensity is >500W / m 2 At this time, the coating changes from hydrophobic to hydrophilic, triggering a capillary heat dissipation mechanism.

[0009] Furthermore, in the metamaterial composite heat collection module, the gradient refractive index nanopillar array is prepared by atomic layer deposition, and the graphene-phase change material composite layer is prepared by hot pressing and uniformly dispersed in the phase change material.

[0010] Furthermore, in the adaptive energy storage and energy management module, the cellular ceramic container is prepared using 3D printing technology, with a hexagonal side length of 1.8mm and a wall thickness of 0.25mm. The bidirectional DC-DC converter adopts synchronous rectification technology and has a switching frequency of 500kHz.

[0011] Furthermore, in the dynamic topology wireless networking module, the blockchain consensus mechanism adopts an improved PoS algorithm, the node weight is determined by the remaining energy and communication contribution, the consensus period is 10 seconds, the ZigBee module is equipped with a power amplifier, and the LoRa module adopts a spreading factor of 7 and a coding rate of 4 / 5.

[0012] Furthermore, in the intelligent power scheduling module, the state space dimension of the deep reinforcement learning algorithm is set to 12 dimensions, covering the current illumination, the illumination prediction for the next hour, the SOC of each node, and the load type. The action space is determined to be an 8-bit PWM signal, and the scheduling period is set to 1 minute.

[0013] Furthermore, in the self-healing photovoltaic unit module, the MEMS micromirror array adopts an electrostatic driving method with a driving voltage range of 0-25V and a power consumption of less than 0.4mW. It is set that when the temperature of the solar cell is detected to be greater than 60°, the micromirrors in the corresponding area deflect 30° to reflect part of the light energy.

[0014] Furthermore, in the environmentally adaptive heat dissipation module, the PNIPAM hydrogel is prepared by free radical polymerization, wherein the crosslinking agent content is 1.5%, and the titanium dioxide nanotube array is prepared by anodic oxidation, with a tube diameter in the range of 80-100 nm and a length in the range of 0.8-1 μm.

[0015] Furthermore, it also includes an intelligent fault diagnosis module: the system deploys vibration sensors and infrared thermal imagers, identifies equipment anomalies through convolutional neural networks, inputs vibration spectrum and thermal image into the diagnostic model, outputs the fault type, and pushes alarm information to the monitoring center through the LoRa network.

[0016] Compared with existing technologies, the beneficial effects of this invention are: The metamaterial composite thermal collector module utilizes a gradient refractive index nanopillar array and graphene phase change material, achieving a daily power generation of 5.8 kWh. Under typical summer weather conditions in Guangzhou, it can maintain continuous power supply for 21 hours, addressing power shortages in remote areas. The adaptive energy storage and management module, through cellular ceramic containers and bidirectional DC-DC conversion technology, improves energy storage efficiency, extending the system's continuous power supply time from 12 hours to 21 hours during Shenzhen's winter, reducing the frequency of diesel generator complementary power supply.

[0017] The dynamic topology wireless networking module constructs a blockchain-based hybrid topology structure, reducing communication latency to 48ms and achieving a throughput of 180TPS for a 50-node network. It supports real-time on-chain recording of device status data, with fault warning latency controlled within 10 seconds. The intelligent power scheduling module utilizes deep reinforcement learning algorithms to improve energy utilization, achieving precise control of load voltage fluctuations within ±3% in laboratory tests, preventing equipment damage due to voltage instability. The self-healing photovoltaic unit and the environmentally adaptive heat dissipation module work together to reduce the cell efficiency degradation rate to 0.3% / year and the phase change material cycle life to 3000 cycles.

[0018] The intelligent fault diagnosis module, through vibration sensors and infrared thermal imagers, provides a 9-second early warning of dangerous conditions such as battery thermal runaway and automatically cuts off the circuit, with a positioning accuracy of 0.5m, reducing accident losses. The system as a whole achieves breakthroughs in power generation efficiency, communication reliability, and intelligent management, and can be widely applied in scenarios such as IoT distributed nodes and microgrids in remote areas, providing a low-power, highly reliable distributed power supply solution for the energy internet and promoting the upgrading of green energy technologies. Attached Figure Description

[0019] Figure 1 This is a schematic block diagram of the distributed solar power supply and wireless networking system proposed in this invention; Figure 2 A comparison chart of energy storage efficiency for different systems under different light intensities; Figure 3 Line chart comparing communication delay distribution in a 50-node network; Figure 4 A bar chart comparing the fault diagnosis accuracy distribution of the traditional system and this system. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In the description of this invention, it should be understood that the terms "center," "longitudinal," "lateral," "length," "width," "thickness," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," "outer," "clockwise," and "counterclockwise," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention.

[0022] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, features defined with "first" and "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. Furthermore, the terms "installed," "connected," and "linked" should be interpreted broadly; for example, they may refer to a fixed connection, a detachable connection, or an integral connection; they may refer to a mechanical connection or an electrical connection; they may refer to a direct connection or an indirect connection through an intermediate medium; and they may refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances. The invention will now be described in further detail with reference to the accompanying drawings.

[0023] Reference Figures 1 to 4 A distributed solar power supply and wireless networking system, comprising the following modules: Metamaterial Composite Thermal Module: First, a gradient refractive index nanopillar array is fabricated on the surface of the solar panel. The bottom silicon pillars are precisely etched onto the silicon wafer surface using reactive ion etching (RIE) to create an array structure with a height of 700 nm and a diameter of 220 nm, with a period controlled at 280 nm. This process requires precise control of parameters such as etching gas flow rate, power, and time to ensure the consistency and accuracy of the silicon pillar structure. Subsequently, atomic layer deposition (ALD) is used to deposit titanium dioxide layer by layer on the bottom silicon pillars. By strictly controlling the type and flow rate of the precursor gas, reaction temperature, and time during the deposition process, the titanium dioxide layer thickness is precisely achieved to 45 nm, forming nanopillars with a diameter of 90 nm and a height of 450 nm.

[0024] Next, a graphene-phase change material composite layer was prepared on the back of the heat collector plate. First, octadecane was mixed with 8 wt% expanded graphite, and ultrasonic dispersion technology was used to ensure thorough and uniform mixing, guaranteeing a uniform distribution of expanded graphite within the octadecane. Then, the mixed phase change material and graphene sheets were alternately stacked and placed in a hot-pressing device. The hot-pressing temperature was set to 50℃ and the pressure to 10 MPa. After a certain period of hot-pressing, a composite layer with a thickness of 2 mm was formed. Tests using a laser thermal conductivity meter showed that its vertical thermal conductivity and other properties met the design requirements, the phase change material's melting point reached approximately 26.3℃, and the latent heat of phase change was 238 kJ / kg.

[0025] For the thermoelectric generators at the edge of the module, Bi2Te3-based material was selected for fabrication. During the fabrication process, parameters such as the material's composition ratio and crystal structure were precisely controlled to ensure power generation performance. The fabricated thermoelectric generators were installed at appropriate positions on the edge of the collector plate. When the temperature difference between the collector plate and the environment reached 15°C, it could stably output a voltage of 3.2V and a power density of 6.2mW / cm². 2 It provides continuous power to the edge computing units.

[0026] Adaptive Energy Storage and Management Module: High-purity alumina ceramic photocurable resin is selected and poured into the resin tank of the photocurable 3D printer. The 3D model of the container is imported into the printing software, and its position and size are calibrated and scaled. Printing parameters are precisely set: exposure time is set to 2-3 seconds per layer, resin layer thickness is 0.05mm, scanning speed is controlled at 2000mm / s, and scanning interval is set to 0.06mm. During printing, a printing platform with low bottom peel force is used to reduce stress deformation. After each layer is printed, the platform rise height is precisely controlled to 0.05mm to ensure the bonding accuracy between layers. After printing, the blank is carefully removed from the printing platform and placed in a UV curing chamber for secondary curing for 30 minutes to enhance the blank's strength. Subsequently, degreasing is performed by placing the blank in a high-temperature furnace and heating it from room temperature to 600℃ at a rate of 5℃ / min, holding it at that temperature for 2 hours to remove organic components from the blank. The temperature was then increased from 600℃ to 1600℃ at a rate of 3℃ / min and held for 4 hours for sintering to densify the green body. After sintering, the green body was allowed to cool naturally to room temperature. The hexagonal pore size of the container was confirmed to be 2.5mm (approximately 1.8mm on each side) and the wall thickness was confirmed to be 0.25mm by laser particle size analyzer and scanning electron microscope.

[0027] Next, a molybdenum disulfide nanocoating was deposited. The sintered ceramic container was placed in the reaction chamber of a chemical vapor deposition (CVD) system. Argon gas was first introduced to purge the reaction chamber for 15 minutes to remove residual air at a flow rate of 200 sccm. The temperature of the reaction chamber was then raised to 650°C at a rate of 5°C / min and maintained stable. A molybdenum source gas (e.g., molybdenum hexacarbonyl) and a sulfur source gas (e.g., hydrogen sulfide) were introduced, with the molybdenum source gas flow rate controlled at 5 sccm and the sulfur source gas flow rate controlled at 10 sccm. Argon gas was simultaneously introduced as a carrier gas at a flow rate of 100 sccm. The reaction time was set to 60 minutes. During the reaction, the coating thickness was precisely controlled by adjusting the gas flow rate and reaction time. After the reaction, the gas source was turned off, but argon gas was continued to be introduced, allowing the reaction chamber to cool to room temperature at a rate of 10°C / min. After removing the container, the coating structure was observed using a transmission electron microscope to ensure that the coating has a layered structure, the interlayer spacing meets the design requirement of 0.62nm, and the coating thickness is precisely 3μm.

[0028] Subsequently, phase change material (PCM) filling was performed. The decanoic acid-lauric acid eutectic PCM was pretreated by placing it in a vacuum drying oven and drying it at 60°C for 24 hours to remove moisture and improve purity. Vacuum impregnation was then used for filling; the ceramic container was placed in a vacuum chamber, and a vacuum of 10⁻⁶ was applied. -3 The pressure was increased to Pa and maintained for 30 minutes to remove air from the pores of the container. The pretreated phase change material was then heated to 40°C until it was molten, and slowly poured into a vacuum chamber. The chamber was maintained under vacuum for 1 hour to allow the phase change material to fully penetrate the pores of the ceramic container. After filling, the vacuum was slowly released, and the container was cooled to room temperature at atmospheric pressure to ensure uniform distribution of the phase change material within the container. Differential scanning calorimetry (DSC) was used to test the filled material, verifying its melting point of 22.5°C and latent heat of phase change of 212 kJ / kg.

[0029] In terms of energy storage system integration, a 50F supercapacitor and a 10Ah lithium battery are rationally connected. A printed circuit board (PCB) is designed, with the mounting positions of the supercapacitor and lithium battery planned to ensure appropriate distance between them and reduce electromagnetic interference. An LM5175 chip is used to construct a bidirectional DC-DC converter. In the peripheral circuit design, suitable power MOSFETs with an on-resistance of less than 10mΩ are selected to reduce switching losses. Input and output filter circuits are designed, using low ESR (equivalent series resistance) ceramic and electrolytic capacitors to ensure circuit stability. The switching frequency is set to 500kHz, and conversion efficiency is improved by optimizing the duty cycle and dead time of the PWM control signal and employing synchronous rectification technology. During PCB fabrication, a four-layer PCB structure is used, with the middle two layers serving as power and ground planes to reduce signal interference. Power circuits and signal circuits are separated during PCB routing to avoid mutual interference. After soldering, the circuit is debugged, and the converter is tested using an electronic load. Under a 2A load, circuit parameters are adjusted to ensure a conversion efficiency of 96.5%-96.8%.

[0030] The energy management algorithm is implemented based on the STM32H750 chip. In the development environment, a new project file is created, and the chip's clock system is configured, setting the main frequency to 480MHz. GPIO pins are initialized, and the ADC channel for acquiring data such as light intensity and battery SOC is configured, setting the sampling frequency to 10Hz. A light intensity acquisition program is written to read the analog signal from the light sensor via the ADC, convert it to a digital value, and calculate the actual light intensity value based on the sensor's calibration curve. A battery SOC calculation program is written, using the ampere-hour integration method combined with the open-circuit voltage method to calculate the battery's SOC value in real time. In the main program, a timer interrupt service function is created to perform data acquisition and algorithm judgment every 100ms (i.e., sampling frequency 10Hz). In the interrupt service function, the current light intensity and battery SOC value are read and compared with a preset threshold (light intensity less than 180W / m²). 2 The system compares the battery's SOC (State of Charge) with the threshold value (less than 35%). When the threshold condition is met, a control signal is output via the GPIO pin to trigger a relay to switch to the mains-powered complementary power supply mode. To ensure system stability, a hysteresis comparison mechanism is added to the algorithm to avoid frequent switching of power supply modes near the threshold. Simultaneously, fault detection and protection functions are implemented in the software, monitoring system parameters such as voltage and current in real time. When abnormal conditions occur, protective measures are taken promptly, such as cutting off the circuit and issuing alarm signals. Under typical winter weather conditions in Shenzhen, the module was installed in an actual solar power supply system for testing, recording the system's power supply duration and various operating parameters. Comparative tests verify that the module can effectively extend the system's continuous power supply duration and improve the system's stability and reliability.

[0031] Dynamic Topology Wireless Networking Module: First, the system adopts a hybrid star-mesh topology. The central node is equipped with both ZigBee and LoRa communication modules. For the ZigBee module, it is set to operate in the 2.4GHz band, with a power amplifier (gain 10dB) installed, and the transmit power adjusted to 20dBm. A 2dBi omnidirectional antenna is used to ensure a receive sensitivity of -100dBm. The LoRa module is set to operate in the 433MHz band, using a spreading factor of 7, a coding rate of 4 / 5, and a transmit power of 22dBm. A 5dBi directional antenna is configured to ensure a receive sensitivity of -137dBm.

[0032] The networking protocol is based on blockchain technology and employs an improved Proof-of-Stake (PoS) consensus algorithm. The PoS consensus algorithm in blockchain originally determined the right to record transactions based on the stake (such as the amount of cryptocurrency held) of a node. However, in this dynamic topology wireless networking module, considering the practical application scenarios of distributed solar power supply and wireless networking systems, the node weights need to better reflect the system's operating status. Therefore, the traditional PoS algorithm is improved by incorporating the node's remaining energy and communication contribution into the weight determination factors. Remaining energy reflects the node's ability to continue working, while communication contribution reflects the node's activity level in data transmission and other communication tasks. This allows for more rational allocation of network resources and improves the stability and efficiency of the network. In the algorithm settings, the node weight is explicitly determined by remaining energy and communication contribution, and the consensus period is set to 10 seconds. To evaluate link quality, a formula is used... The RSSI fluctuations are smoothed using Kalman filtering with filter parameters Q=0.01 and R=0.1. Each node monitors the received signal strength RSSI and its remaining energy Ei in real time. When a node's RSSI is less than -85dBm or its remaining energy is less than 20%, a topology reconfiguration mechanism is triggered.

[0033] In the 50-node test environment setup, node locations were strategically arranged to simulate real-world application scenarios. Network performance was tested, ultimately achieving a network throughput of 180 TPS, a communication latency of 48ms, and a packet loss rate of <0.5%, meeting design requirements. Simultaneously, the system supports self-organizing networking of 50 nodes, ensuring stable communication for both distributed solar power supply and wireless networking systems.

[0034] Intelligent power scheduling module: For the BH1750 light sensor, a suitable installation location was selected in an open area with uniform sunlight around the solar panel to avoid shading affecting measurement accuracy. Internal circuit parameters were adjusted to ensure a measurement range of 0-2000 W / m. 2The DS18B20 temperature sensor is connected to the main control circuit via an SPI interface and a 5Hz sampling frequency, enabling it to capture real-time changes in light intensity. Thermal grease is used to tightly adhere the sensor to the surface of key heat-generating components such as energy storage devices and power conversion modules. It connects to the system via a single-bus communication method and has been calibrated to stabilize its accuracy at ±0.3℃, acquiring ambient and equipment temperature data at a 5Hz frequency. The INA219 current sensor is connected in series in the load circuit strictly according to circuit specifications, establishing a connection with the main control chip via the I2C communication protocol. A 5Hz sampling frequency is set to accurately measure load current within the 0-5A range with an accuracy of ±1%.

[0035] The STM32H750 main control chip was selected. In its development environment, initialization code was written to configure the chip's clock frequency to 480MHz and enable necessary peripheral interfaces such as SPI and I2C. Data structures and functions related to the deep reinforcement learning algorithm (PPO) were defined in the code. For the state space, a program was written to acquire real-time data from the current light intensity sensor, the predicted light intensity for the next hour based on an LSTM model, and to read the SOC value of each energy storage device in real-time (calculated by measuring battery voltage and current and combining it with a battery charge / discharge model), as well as identify the load type (by analyzing the load current waveform characteristics and comparing it with a preset resistive and inductive load feature library), totaling 12 dimensions of data. The action space was defined as an 8-bit PWM signal. The resolution of the PWM signal generation module was set to 0.39% in the program, and a function was written to adjust the duty cycle of the control pins of the DC-DC converter to achieve power distribution.

[0036] The scheduling period is set to 1 minute. A timer interrupt is set in the algorithm program to trigger a power scheduling calculation every minute. The objective function is defined. Calculate the load power consumption P in real time in the code. load (Obtained by multiplying the collected load current and voltage), system energy loss E loss (Calculated based on parameters such as power conversion module efficiency and line resistance), total system energy E total (The sum of the energy stored in the energy storage device and the current input energy). The reward function is designed as follows: The program monitors the SOC change rate ΔSOC of the energy storage unit in real time (calculated by comparing the SOC values ​​at adjacent times). By continuously running the algorithm and utilizing the backpropagation mechanism of the neural network, the parameters in the algorithm are adjusted based on the feedback results of the reward function to optimize the power allocation strategy. At the same time, the load voltage is monitored in real time, and the PWM signal is adjusted through the PID control algorithm to keep the load voltage fluctuation within ±3%.

[0037] This invention also includes a self-healing photovoltaic unit module: silver nitrate is used as a precursor, and silver nanoparticles with a particle size of 15 nm are synthesized by sodium borohydride reduction. Polyvinylpyrrolidone (PVP) is added as a stabilizer during the synthesis process. The reaction temperature is precisely controlled at 40°C, the stirring speed at 300 rpm, and the reaction time at 2 hours to ensure uniform silver nanoparticle size. The synthesized silver nanoparticles are dispersed in a 2 g / L dopamine-Tris buffer solution, with the dopamine content controlled at 25 wt%. Dispersion is performed using an ultrasonic cell disruptor at a power of 200 W for 30 minutes to ensure uniform distribution of the silver nanoparticles in the dopamine solution. The mixed solution is then sprayed onto the surface of the solar panel using an air spraying process. The spray gun pressure is set to 0.3 MPa, the spraying distance is 15 cm, and the spray gun movement speed is 10 cm / s, forming a composite coating with a thickness of 12 μm. After spraying, the solar panel is placed in a 50°C oven for curing for 2 hours to allow the dopamine to fully cross-link and form a stable coating structure.

[0038] For the fabrication of the MEMS micromirror array, SOI (silicon-on-insulator) wafers were used, with a top silicon layer thickness of 2 μm and an intermediate oxide layer thickness of 1 μm. The micromirror structure was defined using photolithography, with AZ9260 photoresist selected and an exposure dose of 200 mJ / cm². 2 The development time was 60 seconds. Deep reactive ion etching (DRIE) was used to etch the silicon layer using SF6 / C4F8 as the etching gas. The etching rate was controlled at 1.5 μm / min, and the etching depth was 2 μm, forming 40 μm × 40 μm micromirror units. A 500 nm thick aluminum reflective layer was deposited on the back of the micromirrors using electron beam evaporation at a rate of 0.5 nm / s and a vacuum level of 5 × 10⁻⁻⁻⁻⁶. 6 Torr fabricated electrostatically driven electrodes with an electrode spacing of 2 μm, achieved through photolithography and aluminum deposition processes.

[0039] During the operation of the photovoltaic unit, a distributed fiber optic temperature sensing system is used for cell temperature monitoring, with a measurement accuracy of ±0.5℃ and a sampling frequency of 1Hz. When the temperature of a cell exceeds 60℃, the temperature sensor transmits the signal to the STM32F407 control chip. The control chip then sends a control signal to the micromirror drive circuit via the SPI bus. The drive circuit uses the HV5812 high-voltage driver chip, which can provide an adjustable drive voltage of 0-25V. It precisely controls the deflection angle of the micromirror using pulse width modulation (PWM) technology. When the temperature exceeds the threshold, the control chip outputs a PWM signal with a 60% duty cycle, causing the micromirror to deflect by 30°, reflecting some light energy and reducing the cell temperature.

[0040] Battery efficiency monitoring employs a four-wire method to measure the cell output current and voltage at a sampling frequency of 5Hz. When a decrease in battery efficiency exceeding 15% is detected in a certain area, the control chip calculates the required micromirror angle adjustment using an algorithm and drives the micromirror array in the corresponding area to adjust its angle. The micromirror angle adjustment utilizes closed-loop control, with a photoelectric position sensor monitoring the micromirror angle in real time and feeding feedback to the control chip for PID regulation, ensuring the micromirror angle is precisely controlled within ±0.5° of the set value.

[0041] For self-healing performance testing, a nanoindenter was used to create scratches 30 μm deep on the coating surface, and the samples were placed in an environmental chamber with 75% humidity and 25°C. The scratch repair process was observed periodically using an optical microscope, with the repair progress recorded every 2 hours. In the 85°C high-temperature aging test, the solar panel was placed in a high-low temperature test chamber, with a temperature cycling range of -40°C to 85°C, each cycle lasting 6 hours, for a total of 1000 cycles. The cell efficiency was tested every 100 cycles, and the efficiency degradation was recorded. These specific process parameters and testing methods ensure the performance and stability of the self-healing photovoltaic module.

[0042] This invention also includes an environmentally adaptive heat dissipation module: In a clean reaction vessel, N-isopropylacrylamide monomer is added, along with N,N'-methylenebisacrylamide crosslinking agent at a precise crosslinking agent content of 1.5%, and an appropriate amount of ammonium persulfate initiator. Deionized water is then added to prepare a homogeneous solution. The reaction system is placed in a 60°C constant-temperature water bath, with a stirring speed of 200 rpm, and reacted for 4 hours. After the reaction, the generated hydrogel solution is uniformly coated onto the surface of the energy storage container, using a scraping method to control the thickness to 150 μm, and then allowed to air dry at room temperature.

[0043] Next, a top-layer titanium dioxide nanotube array was prepared. Using titanium foil as a substrate, it was placed in an ethylene glycol electrolyte, with an appropriate amount of ammonium fluoride added as the electrolyte at a concentration of 0.5 wt%. The titanium foil was used as the anode, and a platinum sheet as the cathode. An 18V DC voltage was applied between the two electrodes, and the reaction was allowed to proceed for 90 minutes. During the reaction, the electrolyte temperature was maintained at 25°C, and the mixture was stirred at 100 rpm using a magnetic stirrer to ensure uniform reaction. After the reaction was complete, the prepared titanium dioxide nanotube array was carefully transferred to the surface of the energy storage container and bonded to the underlying hydrogel.

[0044] The biomimetic siphon pipe was fabricated using 3D printing technology with polylactic acid (PLA) as the material. The 3D printer parameters were set as follows: printing temperature 210℃, printing speed 50mm / s, and layer thickness 0.1mm. Based on the design model, a pipe with an inner diameter of 1mm and a helix angle of 10° was printed. The printed biomimetic siphon pipe was then connected to a micro-peristaltic pump, which was set to a flow rate of 0.5L / min and a power consumption of 0.8W.

[0045] In actual operation, when the temperature sensor detects a temperature exceeding 32℃, or the light sensor detects a light intensity greater than 500W / m², the operation will be affected. 2 At that time, due to the temperature-sensitive properties of the PNIPAM hydrogel and the photoresponse properties of the titanium dioxide nanotube array, the contact angle of the coating decreased from 152° to 4°, initiating capillary heat dissipation. Simultaneously, a micro-peristaltic pump was activated, and the coolant in the biomimetic siphon pipe circulated, carrying away the heat from the energy storage unit and reducing its temperature by 6-8°C. This ensured the stable operation of the energy storage unit in high-temperature environments and extended the cycle life of the phase change material.

[0046] This invention also includes an intelligent fault diagnosis module: the ADXL345 vibration sensor is securely installed in key parts of the equipment to ensure it can accurately sense the vibration during equipment operation; simultaneously, the AMG8833 infrared thermal imager is strategically positioned to fully cover the monitoring area. Both devices are configured to collect data every 10 seconds to ensure the data update frequency meets fault diagnosis requirements.

[0047] For the acquired vibration signals, the Fast Fourier Transform (FFT) algorithm is used to convert the time-domain signals into frequency-domain signals and extract the characteristic frequencies. For example, for common faults in different equipment, the corresponding characteristic frequency ranges are analyzed; for instance, equipment imbalance faults may have obvious characteristics in a specific low-frequency band. For infrared thermal images, bilinear interpolation is used to improve the original resolution to 32×32, enhancing image details for better analysis of temperature distribution.

[0048] On the software side, a CNN diagnostic model was deployed using the TensorFlow Lite framework. When constructing the model structure, the input layer was set to 64×64 pixels to adapt to the size of the processed infrared thermal image. Three convolutional layers were built sequentially, using different kernel sizes and strides to extract depth features from the image; two pooling layers were added to reduce data dimensionality and computational cost; then two fully connected layers were constructed to comprehensively analyze the extracted features. The final output layer was set to represent five fault categories, corresponding to different types of equipment faults. The model was trained using 1000 sets of sample data containing various fault scenarios. During training, the model parameters were continuously adjusted and the loss function optimized to improve the model's accuracy and generalization ability.

[0049] In actual operation, when the model detects that the battery temperature exceeds 70°C and the temperature rise rate is greater than 2°C / min, it is determined to be a battery thermal runaway. At this time, the system immediately triggers the protection mechanism, automatically cutting off the relevant circuits to prevent the fault from worsening. At the same time, alarm information is sent through the LoRa wireless communication module, and the positioning algorithm achieves a positioning accuracy of 0.5m, which facilitates maintenance personnel to quickly find the fault location and carry out timely handling.

[0050] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A distributed solar power supply and wireless networking system, characterized in that, Includes the following modules: Metamaterial composite solar collector module: A gradient refractive index nanopillar array composed of titanium dioxide and silicon is integrated on the surface of a solar panel. The structure is constructed using the principle of multilayer film interference. A phase change material composed of octadecane and expanded graphite is applied to the back of the solar collector. The thermal conductivity is improved by using a graphene network. A thermoelectric generator is embedded at the edge of the module. The temperature difference between the solar collector and the environment is set to >12℃ as the power generation trigger condition. Adaptive energy storage and energy management module: It adopts a honeycomb ceramic thermal storage container, filled with decanoic acid-lauric acid eutectic phase change material, and the inner wall of the container is coated with molybdenum disulfide nano-coating. The energy storage system integrates supercapacitors and lithium batteries, is equipped with a bidirectional DC-DC converter, and designs an energy management algorithm to adjust the charging and discharging strategy according to light intensity and load demand. Dynamic Topology Wireless Networking Module: Constructs a blockchain-based distributed networking protocol. Each node is equipped with both ZigBee and LoRa communication modules. The networking algorithm adjusts the topology by optimizing link quality and energy consumption. The link quality evaluation formula is as follows: , For the link quality between nodes i and j, To receive signal strength, The remaining energy of node i. As initial energy, , These are the weighting coefficients; Intelligent power scheduling module: Deploys light, temperature, and load current sensors; uses an STM32H750 main control chip; runs a deep reinforcement learning algorithm to optimize power allocation; the action space is the output power of each node; the scheduling objective function is: in, For load power, For energy loss, For total energy, These are the weighting coefficients.

2. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, It also includes a self-healing photovoltaic unit module: the surface of the solar panel is coated with a dopamine-silver nanoparticle composite coating, which self-heals through a biomimetic mussel adhesion mechanism. The module has a built-in microelectromechanical system micromirror array. When the battery efficiency in a certain area drops by more than 15%, the micromirrors automatically adjust their angle to focus the light energy.

3. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, It also includes an environmentally adaptive heat dissipation module: a temperature-sensitive and light-sensitive dual-response coating is prepared on the surface of the energy storage container, with a bottom layer of polyisopropylacrylamide hydrogel and a top layer of titanium dioxide nanotubes. When the temperature is >32℃ or the light intensity is >500W / m 2 At this time, the coating changes from hydrophobic to hydrophilic, triggering a capillary heat dissipation mechanism.

4. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, In the metamaterial composite heat collection module, the gradient refractive index nanopillar array is prepared by atomic layer deposition, and the graphene-phase change material composite layer is prepared by hot pressing and uniformly dispersed in the phase change material.

5. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, In the adaptive energy storage and energy management module, the cellular ceramic container is prepared using 3D printing technology, with a hexagonal side length of 1.8mm and a wall thickness of 0.25mm. The bidirectional DC-DC converter adopts synchronous rectification technology and has a switching frequency of 500kHz.

6. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, In the dynamic topology wireless networking module, the blockchain consensus mechanism adopts an improved PoS algorithm, the node weight is determined by the remaining energy and communication contribution, the consensus period is 10 seconds, the ZigBee module is equipped with a power amplifier, and the LoRa module adopts a spreading factor of 7 and a coding rate of 4 / 5.

7. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, In the intelligent power scheduling module, the state space dimension of the deep reinforcement learning algorithm is set to 12 dimensions, covering the current illumination, the illumination prediction for the next hour, the SOC of each node, and the load type. The action space is determined to be an 8-bit PWM signal, and the scheduling period is set to 1 minute.

8. The distributed solar power supply and wireless networking system according to claim 2, characterized in that, In the self-healing photovoltaic unit module, the MEMS micromirror array adopts an electrostatic driving method with a driving voltage range of 0-25V and a power consumption of less than 0.4mW. When the temperature of the solar cell is detected to be greater than 60°, the micromirrors in the corresponding area deflect 30° to reflect part of the light energy.

9. The distributed solar power supply and wireless networking system according to claim 3, characterized in that, In the environmentally adaptive heat dissipation module, the PNIPAM hydrogel is prepared by free radical polymerization, with a crosslinking agent content of 1.5%, and the titanium dioxide nanotube array is prepared by anodic oxidation, with a tube diameter in the range of 80-100 nm and a length in the range of 0.8-1 μm.

10. The distributed solar power supply and wireless networking system according to claim 1, characterized in that, It also includes an intelligent fault diagnosis module: the system deploys vibration sensors and infrared thermal imagers, identifies equipment abnormalities through convolutional neural networks, inputs vibration spectrum and thermal image into the diagnostic model, outputs the fault type, and pushes alarm information to the monitoring center through the LoRa network.

Images