Dynamic adaptive control method for improving photovoltaic power generation efficiency

Abstract

The invention discloses a dynamic adaptive control method for improving photovoltaic power generation efficiency, and the method comprises the steps: monitoring the temperature distribution of a photovoltaic panel in real time through a fiber bragg grating (FBG) sensor, and dynamically adjusting the hydrogel evaporative cooling and thermoelectric generator (TEG) waste heat recovery power in combination with a fuzzy PID algorithm. According to the method, distributed temperature feedback, collaborative cooling-power generation optimization and dual-mode energy management are integrated, so that the temperature of a PV panel is reduced by 15-20 DEG C, the power generation efficiency is improved by 21.5%, and the TEG waste heat recovery power reaches 713 mW / m < 2 >. The device has the advantages of efficient cooling, waste heat utilization, service life prolonging and the like, and is suitable for large-scale photovoltaic power stations in high-temperature areas.

Description

Technical Field

[0001] The present invention relates to the field of photovoltaic power generation technology, and in particular to a collaborative control method based on real-time temperature monitoring and dynamic feedback, which is used to improve photovoltaic power generation efficiency and extend component life. Background Art

[0002] In actual operation, photovoltaic modules absorb a large amount of solar radiation and generate high temperatures, resulting in a significant reduction in power generation efficiency. Studies have shown that crystalline silicon photovoltaic modules can only convert about 6%-25% of the incident light energy into electrical energy, and the remaining energy accumulates on the panel surface in the form of heat energy, making its operating temperature 20-40°C higher than the ambient temperature. For every 1°C increase in temperature, the efficiency of photovoltaic cells decreases by 0.4%-0.5%. In high-irradiation areas such as deserts, the panel temperature can reach over 70°C, and the power generation loss exceeds 15%. In addition, long-term high-temperature environments will accelerate the aging of packaging materials and the corrosion of battery grid lines, resulting in a 20%-30% shortening of module life, which seriously restricts the economy and reliability of photovoltaic power stations.

[0003] Current photovoltaic thermal management technologies mainly include active cooling and passive cooling. Active cooling technologies such as air cooling and water cooling rely on external energy to drive the circulation of cooling media. Although they can achieve effective cooling, they have problems such as high energy consumption, waste of water resources and complex systems. For example, spray cooling requires continuous water supply, which is limited in its application in arid areas and is prone to mineral deposition; the energy consumption of semiconductor refrigeration is unbalanced with the cooling benefits, and the actual energy efficiency ratio (COP) is less than 0.8. Although passive cooling technology does not require additional energy consumption, its performance is limited. For example, radiation cooling materials (such as SiO 2 The coating is limited by the transmittance of the atmospheric window, and the maximum cooling power is only 120W / m 2 ; Phase change materials (PCM) have low thermal conductivity (<0.3W / m·K) and regenerate slowly after phase change; although hygroscopic evaporative materials (such as LiCl hydrogel) can achieve cooling by absorbing moisture at night and evaporating during the day, their cycle stability is poor (performance decays by 40% after 50 cycles), and they are not effectively coordinated with waste heat recovery technology, and more than 80% of the heat energy is still lost to the atmosphere.

[0004] Existing technologies also have the defect of disconnection between monitoring and control. Traditional temperature detection methods are difficult to meet the needs of accurate and distributed monitoring. For example, thermocouples need to be installed destructively and only provide single-point data. Infrared thermal imagers are interfered by environmental reflections, and the outdoor error reaches ±3°C. In addition, due to the lack of a dynamic control mechanism based on real-time data, the cooling and power generation links operate in isolation, and the overall energy efficiency improvement is limited. Therefore, it is urgent to develop a solution that integrates efficient cooling, waste heat recovery and intelligent control, and break through the bottleneck of photovoltaic thermal management technology through the collaborative innovation of materials, structures and algorithms. Summary of the invention

[0005] The present invention provides a dynamic adaptive control method for improving the photovoltaic power generation efficiency, aiming to solve the problems of low cooling efficiency, poor waste heat utilization rate and disconnection of monitoring and control in the existing photovoltaic thermal management technology. The method realizes the closed-loop optimization of photovoltaic module temperature control and energy recovery through the multi-dimensional coordination of hydrogel evaporation cooling, thermoelectric waste heat recovery and optical fiber sensing technology. The specific scheme is as follows:

[0006] (1) Multilayer composite thermal management module

[0007] The thermal management module is composed of a thermoelectric generator layer and a hydrogel composite layer integrated in sequence on the back of the photovoltaic module.

[0008] Preferably, the Bi 2 Te 3 / Sb 2 Te 3 heterojunction thermoelectric unit array, the unit size is 20mm×20mm×3mm, and the Seebeck coefficient is ≥200μV / K.

[0009] Preferably, the cold end of the TEG is coupled with the hydrogel layer through thermal conductive silicone grease (thermal conductivity ≥3W / m·K), and the hot end is directly attached to the photovoltaic backplane, and the contact thermal resistance ≤0.01K·m 2 / W to ensure efficient heat conduction.

[0010] Preferably, the LiCl / Na 2 SiO 3 / PVA-PAm double-network hydrogel, and its preparation method includes:

[0011] (1) Blend polyvinyl alcohol (PVA) and acrylamide (AM) at a molar ratio of 1:3, and add 12wt% Na 2 SiO 3 solution to form a primary network;

[0012] (2) Immerse in 35wt% LiCl solution for 24 hours to construct a double-network structure through ionic crosslinking;

[0013] (3) Regulate the pore size distribution through freeze-thaw cycles, the average pore size is 2μm, and the specific surface area reaches 150m 2 / g.

[0014] Preferably, the infrared emissivity of the hydrogel is ≥0.98 (8-13μm band), the visible light transmittance is ≥85%, the evaporation cooling power reaches 332W / m 2 (under 1000W / m 2 irradiation), and the nocturnal moisture absorption capacity is ≥22.5g / g (90%RH environment).

[0015] (2) Distributed temperature feedback system

[0016] The system uses a fiber Bragg grating (FBG) sensor array to achieve high-precision temperature monitoring.

[0017] Preferably, the 2×6 grid layout covers the surface of a 245mm×145mm photovoltaic panel, with a spacing of 30mm between adjacent sensors and a spatial resolution of 20mm×20mm.

[0018] Preferably, the FBG sensors are fixed by ultraviolet curable glue, the sampling frequency of the wavelength demodulator is ≥100Hz, and the wavelength resolution is ≤1pm.

[0019] Preferably, the built-in temperature self-calibration module eliminates environmental disturbance errors through a reference grating. Combining with an improved Kalman filter algorithm, the temperature monitoring accuracy is improved to ±0.1°C, local temperature gradients of <1°C can be identified, and a thermal map is generated in real time to locate the hot spot area.

[0020] (III) Adaptive cooperative control strategy

[0021] The strategy realizes the dynamic optimization of cooling and power generation based on the fuzzy PID algorithm, specifically including:

[0022] (1) Control logic

[0023] ① The input variables include temperature deviation (e), deviation change rate (Δe), light intensity (G), and relative humidity (RH), where e and Δe are divided into 7 and 5 fuzzy sets respectively;

[0024] ② The fuzzy rule base contains 36 IF-THEN rules, for example: "If e = positive large and Δe = positive fast, then the output hydrogel flow rate = maximum, and the TEG load resistance = minimum";

[0025] c. Defuzzification uses the centroid method, and the control response time is ≤200ms.

[0026] (2) TEG waste heat recovery

[0027] The TEG load resistance is dynamically adjusted by the maximum power point tracking (MPPT) algorithm, and when ΔT≥8°C, the waste heat recovery power density is ≥713mW / m 2 , and the energy conversion efficiency is ≥5%.

[0028] (3) Dual-mode energy management

[0029] ① Mode 1 (normal operation): Photovoltaic power is connected to the grid through an inverter, and TEG power drives a micro pump (≤2W) and a sensor network;

[0030] ② Mode 2 (extreme environment): When the humidity ≥80% and the light intensity ≤200W / m 2When switching to the off-grid mode, the TEG power is stored in the lithium battery-supercapacitor hybrid energy storage system (charge and discharge efficiency ≥ 95%).

[0031] One technical solution provided in the embodiment of the present application has at least the following technical effects or advantages:

[0032] (1) The temperature of the photovoltaic panel is reduced by 15 - 20 °C (the peak temperature difference reaches 25 °C), the power generation efficiency is increased by 21.5% - 24%, and the annual power generation of a 100 MW power station is increased by 18.7%;

[0033] (2) The waste heat recovery power density of the thermoelectric generator can reach 713 mW / m 2 , meeting 60% of the energy consumption requirements of the off-grid equipment of the power station;

[0034] (3) The standard deviation of the temperature fluctuation is reduced by 76%, the risk of packaging failure caused by thermal stress is decreased by 40%, and the component life is extended by 3 - 5 years;

[0035] (4) The annual income of a 100 MW power station increases by $580,000, and the CO 2 emission reduction is 19,800 tons, which is equivalent to the annual emissions of 5,500 fuel vehicles. Description of the Drawings

[0036] Figure 1 It is a schematic diagram of the system structure of a photovoltaic module according to Embodiment 1 of the present application;

[0037] 1 - polysilicon PV panel, 2 - LNP hydrogel layer (including radiation enhancement layer and evaporation microchannel), 3 - TEG array (cold end contacts the hydrogel, hot end adheres to the PV backplane), 4 - FBG sensor network; 5 - edge computing control box (including STM32H7 main control and LoRa communication module), 6 - dual-mode energy storage system (supercapacitor + lithium battery).

[0038] Figure 2 It is a dynamic control flow chart of photovoltaic power generation according to Embodiment 2 of the present application. Detailed Embodiments

[0039] The following further describes the present invention in detail with reference to embodiments, but the protection scope of the present invention is not limited to the described embodiments.

[0040] Embodiment 1

[0041] This embodiment provides a single-component level implementation method, which specifically includes the following four steps:

[0042] (1) Integration of the thermal management module

[0043] In this embodiment, the photovoltaic module selects a 245 mm × 145 mm polysilicon panel (conversion efficiency 11.9%), and the following structures are sequentially integrated on the back:

[0044] (1) Use 15 Bi 2 Te 3 -based thermoelectric units (size 20 mm × 20 mm × 3 mm), connected in series by serpentine wiring. The cold ends are coated with high thermal conductivity silicone grease (Shin-Etsu X-23-7783D, thermal conductivity 3.5 W / m·K) and closely attached to the hydrogel layer;

[0045] (2) Prepare the LiCl / Na 2 SiO 3 / PVA-PAm double-network hydrogel according to the following steps:

[0046] ① Dissolve 4.2 g of acrylamide (AM) and 1.8 g of polyvinyl alcohol (PVA) in 40 mL of deionized water, add 5 mL of 12 wt% Na 2 SiO 3 solution, and stir magnetically for 30 minutes;

[0047] ② Add 0.15 g of N,N′-methylenebisacrylamide (MBA) crosslinking agent and 20 μL of N,N,N′,N′-tetramethylethylenediamine (TEMED), stir for 5 minutes and then inject into the mold, and cure at 80 °C for 1 hour to obtain the primary gel;

[0048] ③ Immerse the primary gel in 35 wt% LiCl solution for 24 hours to complete ionic crosslinking, and the final thickness is 2.0 ± 0.1 mm.

[0049] (3) Use 6061 aluminum alloy plate (thickness 1.5 mm, thermal conductivity 167 W / m·K), and perform surface anodization treatment to enhance weather resistance.

[0050] (II) Sensor network deployment

[0051] The FBG sensor array in this embodiment is installed in the following manner:

[0052] (1) Fix 12 FBG sensors (central wavelength 1550 ± 1 nm) on the PV surface according to a 2×6 grid, with a lateral spacing of 30 mm and a longitudinal spacing of 25 mm between adjacent sensors;

[0053] (2) Use ultraviolet curable glue (LOCTITE 352) to adhere the sensors to the panel surface, avoiding blocking the solar cells. Calibrate the sensors through a constant temperature water bath (25 ± 0.1 °C), with a sensitivity of 10.07 pm / °C and a linearity R 2 = 0.999;

[0054] (3) Adopt a MOI sm125 demodulator, with a sampling frequency of 1 kHz and a wavelength resolution of 0.5 pm, and transmit it to the control unit through the RS485 interface.

[0055] (3) Control System Configuration

[0056] The hardware and software design of the control unit in this embodiment is as follows:

[0057] (1) Hardware Architecture

[0058] ① Main control chip: STM32H743VIT6 (ARM Cortex-M7 core, main frequency 480MHz);

[0059] ② Actuator: Piezoelectric ceramic micropump (model SMP-300, flow range 0 - 50 μL / min), digital potentiometer (AD5272, resistance range 0 - 100 Ω);

[0060] ③ Communication module: LoRa wireless transmission (frequency band 433MHz, transmission distance ≥ 1km).

[0061] (2) Software Algorithm

[0062] ① Fuzzy PID controller: Deployed in the FreeRTOS real-time system, with input variables e (-10 °C to +10 °C), Δe (-2 °C / min to +2 °C / min), and output variables being the duty cycle of the micropump (0% - 100%) and the TEG load resistance value;

[0063] ② Thermogram generation: Reconstruct the temperature field based on the Inverse Distance Weighting (IDW) algorithm, with a refresh rate of 10Hz.

[0064] (4) Outdoor Performance Test

[0065] The test conditions and results in this embodiment are as follows:

[0066] (1) Test environment: Noon in summer in Nanjing (solar irradiance 1000W / m 2 , ambient temperature 35 °C, relative humidity 60%, wind speed 1.2m / s);

[0067] (2) Temperature control: The average temperature of the PV surface is stabilized at 42 ± 0.5 °C (control group 62 ± 1.2 °C), and the local hot spot temperature difference ≤ 1.5 °C;

[0068] (3) Power generation performance: The maximum power is increased from 0.19W to 0.235W (+23.7%), and the fill factor is increased from 0.601 to 0.644. The open circuit voltage is 0.51V, and the maximum power is 713mW / m 2 , which can drive a 2W LED to work continuously for 4.2 hours;

[0069] (4) Cycle stability: After continuous operation for 30 days, the moisture absorption of the hydrogel decays ≤ 3%, and the output power fluctuation of the TEG ≤ 5%.

[0070] Example 2

[0071] This embodiment provides a power station-level deployment method, and the system architecture and economic verification are as follows.

[0072] (1) System Architecture

[0073] The 100MW photovoltaic power station in this embodiment is transformed according to the following plan:

[0074] (1) Unit division: Each 1MW sub-array is configured with 1 central controller (NVIDIA Jetson AGX Orin, computing power 32 TOPS), and each 20 components form an independent control unit;

[0075] (2) TEG power management: A flow battery (rated capacity 50kWh, efficiency 78%) is used to store TEG power for monitoring cameras and cleaning robots. The DC-DC converter (efficiency ≥92%) realizes MPPT control, and the maximum power point voltage tracking error ≤2%;

[0076] (3) Communication network: Networking is carried out through a LoRaWAN gateway, the data upload interval is 10s, and the packet loss rate ≤0.1%.

[0077] (2) Economic Verification

[0078] The economic analysis in this embodiment is based on the electricity price in 2023 ($0.12 / kWh) and the carbon trading price ($50 / ton CO 2 ):

[0079] (1) Incremental cost: The component transformation cost is $0.12 / W (including materials, installation and commissioning), and the total investment in the 100MW power station is $12 million;

[0080] (2) Revenue calculation: The annual power generation increases from 135GWh to 160GWh (+18.5%), and the additional income is $3 million. The annual carbon emission reduction is 19,800 tons of CO 2 , and the carbon trading income is $990,000. Through static calculation, the payback period is 2.8 years;

[0081] (3) Operation and maintenance cost: The hydrogel is replaced every 5 years, the service life of the TEG module ≥10 years, and the average annual maintenance cost ratio ≤1.5%.

[0082] Example 3

[0083] This embodiment provides an adaptation method in extreme environments.

[0084] (1) Implementation Scheme

[0085] This embodiment is optimized for the Neom Desert in Saudi Arabia (daytime peak temperature of 50 °C, nighttime humidity ≤ 15%).

[0086] (1) Hydrogel strengthening: Increase the LiCl concentration to 40 wt%, and add 0.5 wt% sodium alginate to enhance the mechanical strength (tensile modulus increased from 701 Pa to 950 Pa);

[0087] (2) Sand prevention design: Install a porous aluminum alloy filter (porosity 85%, pore diameter 200 μm) outside the heat dissipation substrate, and automatically blow and remove dust every day;

[0088] (3) Energy storage expansion: Use a phase change heat storage material (paraffin / expanded graphite composite, latent heat 180 J / g) to buffer the nighttime temperature fluctuations.

[0089] (2) Test results

[0090] After continuous operation for 6 months using this method, the specific results are as follows:

[0091] (1) Temperature control: The daytime peak temperature of the PV panel ≤ 55 °C (75 °C for the control group), and the nighttime moisture absorption is stable at 2.8 g / g;

[0092] (2) Power generation performance: The photovoltaic efficiency is maintained at 11.2% (the control group decays to 9.5%); the TEG output power is increased to 1.1 W / m 2 (ΔT = 12 °C);

[0093] (3) Durability: The dust accumulation on the filter ≤ 5 g / m 2 , and there is no salt precipitation or cracking in the hydrogel.

Claims

1. A dynamic adaptive control method for improving photovoltaic power generation efficiency, characterized in that: The following steps are involved: ① Real-time collection of photovoltaic (PV) panel surface temperature field data through a distributed fiber Bragg grating (FBG) sensor array, and simultaneous acquisition of environmental parameters (including light intensity, ambient temperature, relative humidity and wind speed); ②Build a PV panel thermodynamic model based on temperature field data, and predict the local hot spot location and temperature gradient distribution through finite element simulation; ③Using fuzzy proportional-integral-differential (PID) control algorithm, dynamically adjust the hydrogel evaporative cooling rate and the waste heat recovery power of the thermoelectric generator (TEG), including: Method 1 generates a flow rate control signal (0-50 μL / min) of a hydrogel microfluidic pump according to the temperature deviation (e) and the deviation change rate (Δe); Method 2 optimizes the load resistance (0-100Ω) in real time based on the temperature difference between the hot and cold ends of the TEG (ΔT≥5°C) to maximize the waste heat recovery power; ④ Implement a dual-mode energy management strategy, including: Mode 1: PV power is fully connected to the grid, and TEG power drives the local cooling system and monitoring equipment; Mode 2 switches to off-grid mode in extreme weather conditions, with TEG energy storage giving priority to maintaining self-sustaining operation of the system.

2. The method according to claim 1, characterized in that The hydrogel is a lithium salt / sodium silicate / polyvinyl alcohol-polyacrylamide (LiCl / Na2SiO3 / PVA-PAm) double network hydrogel, and its technical parameters include: ① Infrared band (8-13μm) emissivity ≥ 0.98, visible light transmittance ≥ 85%; ②Evaporative cooling power ≥300W / m 2 (1000W / m 2 irradiance), continuous working time during the day ≥ 8h; ③ Nighttime moisture absorption and regeneration capacity ≥2.5g / g (90% RH environment), cycle stability ≥500 times without performance degradation.

3. The method according to claim 2, characterized in that The preparation method of the hydrogel comprises the following steps: ① Blend polyvinyl alcohol (PVA) and acrylamide (AM) at a molar ratio of 1:2-1:4, and add 10%-15% by mass fraction of Na2SiO3 solution to form a primary gel network; ② Immerse the primary gel in a 30%-40% LiCl solution for 24 hours to construct a double network structure through an ion replacement reaction; ③ The freeze-thaw method was used to regulate the pore size distribution of the gel (average pore size 1-3 μm), and a radiation-enhanced microstructure was formed through a surface imprinting process.

4. The method according to claim 1, characterized in that The thermoelectric generator (TEG) module uses Bi2Te3 / Sb2Te3 heterojunction thermoelectric material, and its technical parameters include: ① Seebeck coefficient ≥ 200μV / K, thermal conductivity ≤ 1.5W / m·K; ②The cold end is directly coupled to the hydrogel layer through thermal grease, and the hot end is attached to the PV backplane (contact thermal resistance ≤ 0.01K·m 2 / W); ③Waste heat recovery power density ≥ 700mW / m 2 (When ΔT≥8℃), the maximum energy conversion efficiency is ≥5%.

5. The method according to claim 1, characterized in that The deployment of the fiber Bragg grating (FBG) sensor array satisfies: ① Cover the PV panel surface with a 2×6 grid distributed layout (spatial resolution 20mm×20mm); ② Use wavelength demodulator to achieve multi-channel synchronous acquisition (sampling frequency ≥ 100Hz, wavelength resolution ≤ 1pm); ③ Built-in temperature self-calibration module, which eliminates environmental disturbance errors through reference grating, and the monitoring accuracy reaches ±0.1℃.

6. The method according to claim 1, characterized in that The implementation of the fuzzy PID control algorithm includes: ① Fuzzification of input variables: temperature deviation (e) is divided into seven fuzzy sets {NB, NM, NS, ZO, PS, PM, PB}, and deviation change rate (Δe) is divided into five fuzzy sets {negative fast, negative slow, zero, positive slow, positive fast}; ② The fuzzy rule base contains 49 IF-THEN rules, for example: if e = PB and Δe = positive fast, then the output hydrogel flow rate = maximum, TEG load resistance = minimum; ③ The defuzzification adopts the center of gravity method to output precise control quantity, and the response time is ≤200ms.

7. The method according to claim 1, characterized in that The dual-mode energy management strategy further includes: ①When the ambient humidity is ≥80% and the light intensity is ≤200W / m 2 When the off-grid mode is automatically enabled; ② The energy storage system adopts a hybrid architecture of lithium-ion batteries and supercapacitors (capacity ratio 3:1), with a charge and discharge efficiency of ≥95%; ③ The grid-connected interface supports the IEEE 1547 standard, and the harmonic distortion rate (THD) is ≤3%.

8. A system for implementing the method of claims 1-7, characterized in that: Includes the following modules: Module: Photovoltaic-hydrogel-TEG integrated unit ① The PV panel (monocrystalline silicon or polycrystalline silicon, efficiency ≥18%) is laminated with the hydrogel layer through a weather-resistant adhesive (peel strength ≥10N / cm); ②The TEG array is connected with serpentine wiring, and the output end is connected to a DC-DC converter (efficiency ≥ 92%); Module 2 Intelligent monitoring and control unit ①Edge computing node (equipped with ARM Cortex-A72 processor, computing power ≥ 2 TOPS); ②Multi-protocol communication module (supports LoRaWAN, NB-IoT and 5G redundant backup); Module 3 Energy Management Unit ①Dual-input MPPT controller (PV and TEG are optimized independently); ②Off-grid load interface (rated power ≤5kW, including overvoltage / undervoltage protection).

9. The system according to claim 8, characterized in that The surface of the hydrogel layer is provided with a bionic microstructure, including: ① Radiation enhancement zone: Sub-wavelength grating (period 2-5μm) is formed by laser etching to improve mid-infrared emissivity; ② Evaporation channel: Use femtosecond laser to process multi-level fractal flow channels (width 50-200μm) to accelerate water transport; ③ Dust-proof coating: Spray hydrophobic silica nanoparticles (contact angle ≥ 150°) to inhibit dust deposition.

10. The system according to claim 8, characterized in that The deployment method of the system includes: Method 1: Compatibility transformation of existing photovoltaic power stations ① Open a hole in the PV back panel to install the TEG module (aperture error ≤ 0.1mm); ② Use robots to automatically coat the hydrogel layer (thickness uniformity ≥ 95%); Method 2: Adaptive design for desert / high humidity areas ①Add anti-wind and sand filter (porosity ≥ 80%, pressure drop ≤ 50Pa); ②Equipped with a dehumidification regeneration device (dew point temperature ≤ -10°C) to ensure the hydrogel's moisture absorption performance at night.

Images