High-voltage-resistant photoelectric conversion module suitable for submarine optical communication and data transmission system

Abstract

The invention provides a high-voltage-resistant photoelectric conversion module suitable for submarine optical communication and a data transmission system, and belongs to the technical field of deep sea optical communication. Aiming at the problems of performance drift, high error rate, difficulty in fault identification and poor link adaptability of a traditional photoelectric module in a deep sea environment with high pressure of 30-100MPa and temperature change of-40-80 DEG C, the invention provides a method for evaluating the health degree of a deep sea through a temperature-pressure coupling dynamic compensation technology, a distortion grading optimization algorithm, a multi-dimensional health degree evaluation model and system-level link adaptive adjustment. 10-60Gbps high-speed transmission is realized, the fault identification accuracy is greater than or equal to 98%, and the method can be widely applied to scenes such as deep-sea oil-gas exploration, a seabed observation network and cross-ocean communication relay, and has the remarkable advantages of high stability, low operation and maintenance cost and wide adaptability. Multiple scenes such as deep-sea oil-gas exploration, a seabed observation network and polar scientific investigation are covered, the transmission rate is 10-60 Gbps, different service requirements can be flexibly met, and remarkable economic and social benefits are achieved.

Description

Technical Field

[0001] This invention belongs to the field of deep-sea optical communication technology, and more specifically, it relates to a high-voltage photoelectric conversion module and data transmission system suitable for submarine optical communication. Background Technology

[0002] Deep-sea communication faces three core challenges: high voltage, drastic temperature changes, and long-distance transmission loss. Traditional optoelectronic modules are prone to output voltage and optical power drift in deep-sea high-pressure (≥30MPa) and wide-temperature (-40-80℃) environments, leading to signal distortion (total harmonic distortion THD > 0.5%) and a surge in bit error rate (≥10⁻¹). 0 ); Fault identification relies on a single parameter (such as output voltage), has a high false positive rate (≥15%), and lacks a comprehensive assessment of the module's health status, leading to a sharp increase in deep-sea maintenance costs. In long-distance transmission, dispersion increases non-linearly with distance and speed. Traditional fixed dispersion compensation schemes deteriorate to the point of being unusable at 60Gbps rates. At the same time, bandwidth equalization strategies cannot meet the low-latency requirements of high-priority services (such as real-time video).

[0003] In summary, existing technologies are insufficient to meet the demands for high-speed, stable, and intelligent optical communication in deep-sea environments, necessitating a comprehensive end-to-end optimized solution. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a high-voltage photoelectric conversion module and data transmission system suitable for submarine optical communication. It solves the technical problems of performance drift, high bit error rate, inaccurate fault identification, and poor adaptability of dispersion and bandwidth allocation in traditional photoelectric modules under deep-sea high-voltage (30-100MPa) and temperature-varying (-40-80℃) environments.

[0005] A high-voltage photoelectric conversion module suitable for submarine optical communication, including a high-voltage adapter unit, a photoelectric conversion core unit, a signal optimization unit, and a temperature and pressure coordinated control unit; The high-pressure adapter unit is adapted to seabed pressure environments of 30-100MPa, with an input voltage of 24-48VDC. The output voltage is adjusted by a self-developed high-pressure compensation formula: U_out=U_in×[1+k1×(P-0.1)+k2×|T-25|], where k1 is the pressure coefficient (0.0015-0.0025 / MPa), k2 is the temperature coefficient (0.0008-0.0012 / ℃), P is the real-time pressure (MPa), T is the real-time temperature (℃), and the output ripple is ≤40mV. The photoelectric conversion core unit adopts an InGaAsPIN and GaN driver chip architecture, with a responsivity ≥0.9A / W and dark current ≤0.7nA at a wavelength of 1550nm. The bias current is optimized by a self-created calibration formula: I_b=I_b0×[1+0.0003×(P-0.1)-0.0002×(T-25)], where I_b0 is the reference bias current; The signal optimization unit has a balanced bandwidth of 1-50GHz, and the control unit has a response time of ≤35μs.

[0006] Preferably, the signal optimization unit adopts a distortion graded optimization algorithm: c relay gain 35-45dB.

[0007] Preferably, the shore-based main control unit includes a heterogeneous data processing subunit, an intelligent power supply management subunit, and a fault location subunit; The heterogeneous processing subunit adopts an FPGA+GPU architecture, with a data cache capacity of ≥1.5GB and a processing latency of ≤6μs; The power supply management subunit outputs 3500-6000VDC, with ripple ≤70mV and power supply efficiency ≥92%; The fault location subunit combines the module H value with OTDR data, achieving a location accuracy of ≤20m and a diagnostic accuracy of ≥98%.

[0008] Preferably, the data transmission steps are as follows: 1) The shore-based main control unit sends out initialization parameters, setting module I_b to I_b0 and relay gain to 40dB; 2) The transmitting module converts the electrical signal into a 10-15dBm optical signal and injects it into the transmission link; 3) The intelligent relay unit collects L_meas and T, P in real time, and amplifies them after compensation according to the L_comp formula; 4) The receiving module restores the electrical signal and processes it through the signal optimization unit; 5) BER monitoring by shore-based units; when BER > 10 -14 At that time, according to ΔG=0.4×(BER / 10 -14 Adjust the relay gain; response time ≤ 12μs.

[0009] Preferably, it also includes an intelligent redundancy unit, which employs a switching decision formula: S=0.4×(ΔP / ΔP0)+0.3×(BER / 10 -14 )+0.2×(1-H)+0.1×(ΔT / ΔT0), where ΔP is the optical power fluctuation and ΔT is the temperature fluctuation. When S≥0.65, the system automatically switches to the backup module / link with a switching time ≤6ms and an interruption time ≤30μs.

[0010] Compared with the prior art, the present invention has the following beneficial effects: Through dynamic compensation via temperature and pressure coupling, the output voltage ripple is ≤40mV and the optical power fluctuation is ≤±0.05dBm, which improves stability compared to traditional solutions. It can work reliably for a long time in a wide temperature environment with high voltage of 30-100MPa and temperature range of -40-80℃. Distortion classification optimization achieves THD ≤ 0.25% and SNR ≥ 58dB. Combined with dynamic dispersion compensation, the bit error rate for 60Gbps transmission is ≤ 10. -14 This reduces costs by 4-5 orders of magnitude compared to traditional solutions; The multi-dimensional health model has a fault identification accuracy of ≥98% and a false alarm rate of ≤1.2%, enabling early warning of faults. Combined with dynamic bandwidth allocation, it improves the efficiency of deep-sea maintenance. It covers multiple scenarios such as deep-sea oil and gas exploration, seabed observation network, and polar scientific research. With a transmission rate of 10-60Gbps, it can flexibly adapt to different business needs and has significant economic and social benefits. Attached Figure Description

[0011] Figure 1 This is the code logic diagram of the distortion grading optimization algorithm in this invention; Figure 2 This is a flowchart of the distortion grading optimization algorithm in this invention; Figure 3 This is a block diagram of the internal structure of the high-voltage photoelectric conversion module in this invention; Figure 4 This is a diagram of the overall architecture of the data transmission system in this invention; Figure 5 This is a circuit architecture diagram of the high-voltage adapter unit in this invention; Figure 6 This is a flowchart of the health assessment and early warning mechanism in this invention; Figure 7 This is a flowchart of the closed-loop control of the intelligent relay unit in this invention. Detailed Implementation

[0012] The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and should not be construed as limiting the scope of the invention.

[0013] Please see Figures 1-7 This invention provides a high-voltage resistant photoelectric conversion module and data transmission system suitable for submarine optical communication. The technical solution of this invention consists of two core components: a high-voltage resistant photoelectric conversion module (hereinafter referred to as the module) and a submarine optical communication data transmission system (hereinafter referred to as the system). The module adopts a four-layer architecture of high-voltage adaptation, photoelectric conversion, signal optimization, and temperature and pressure coordinated control, and achieves dynamic matching between environment and performance through four self-created formulas. The system adopts a link architecture of shore-based master control - intelligent relay - module node - redundancy backup, and achieves adaptive adjustment of the entire link through three self-created formulas.

[0014] This invention can be summarized as follows: A temperature-pressure coupling compensation mechanism is proposed, which uses voltage and current dual formula calibration to control the module output ripple within 40mV; A distortion-grading optimization algorithm was designed to improve the signal-to-noise ratio to over 58dB. A multi-dimensional health model was constructed, achieving a fault identification accuracy rate of ≥98%. Achieve system-level dynamic link compensation with a bit error rate of ≤10 at 60Gbps transmission speed. -14 .

[0015] Specific implementation of the high-voltage photoelectric conversion module: Module Hardware Overview and Component Selection: The module adopts an integrated solution of ceramic substrate + titanium alloy packaging. The titanium alloy shell is model TC4 with a wall thickness of 12mm (verified by finite element analysis to show no deformation under 100MPa). The interior is filled with thermally conductive and insulating silicone (Dow Corning SE4420, dielectric strength 20kV / mm). The component selection for each core unit has undergone five rounds of high-voltage environmental testing to ensure parameter stability. Specific selections are shown in the table below: Core Unit Key components Model Specifications Core parameters (25℃ / 0.1MPa) High-voltage compatibility verification (100MPa) High voltage adapter unit isolated power supply TIDCP010505BP Input 24-48V, output 5V, isolation 5.5kVrms Output ripple 38mV, efficiency 86% High voltage adapter unit Surge protection MOV-10D561K+GDT-2R5G Surge current 15kA (8 / 20μs), response 6ns Rated output returns to normal within 100μs after surge. Photoelectric conversion core IngaAsPIN HAMAMATSUG8371-01 1550nm responsivity 0.92A / W, dark current 0.6nA The responsivity decreases by ≤0.02A / W, and the dark current increases to 0.7nA. Photoelectric conversion core GaN driver chip CreeCGD15000 Drive current 0-1A, switching frequency 100MHz Current fluctuation ≤ ±1mA Signal optimization unit Adaptive Equalizer ADIAD9250 Bandwidth 1-50GHz, supports 8 / 16th order equalization Equalization accuracy decrease ≤0.1dB Temperature and pressure coordinated control Adaptive Equalizer STM32H743VIT6 400MHz main frequency, 16-bit AD sampling Sampling error ≤ ±0.5% Temperature and pressure coordinated control Temperature and pressure sensor Keller33X+AD590 Pressure: 0-120 MPa (±0.1 MPa); Temperature: -55-150℃ (±0.5℃) Pressure error ±0.2MPa, temperature error ±0.8℃ High voltage adapter unit: The high-voltage adapter unit is the core power source for the module to withstand the high voltage of the deep sea. Its function is to convert the 24-48VDC input from the shore base into a stable 5VDC to power other units. Its core is the high-voltage compensation formula, which solves the problem of excessive ripple caused by changes in the dielectric constant of capacitors under high voltage in traditional fixed-voltage output.

[0016] Circuit structure design: The circuit adopts a four-level architecture: rectification and filtering, isolation and conversion, feedback regulation, and surge protection. Rectifier and filter stage: A π-type filter is formed by a 470μF / 100V electrolytic capacitor and a 0.1μF ceramic capacitor to filter out high-frequency noise from the input; Isolation conversion stage: TIDCP010505BP achieves high voltage isolation to prevent high voltage from the deep sea from being conducted to subsequent circuits; Feedback regulation stage: ADIOP07 operational amplifier + TIDAC8552 DAC chip, receiving signals from the control unit to regulate the output voltage; Surge protection level: MOV varistor in parallel with GDT gas discharge tube to suppress surge impacts introduced by the cable (such as instantaneous high voltage caused by fishing boat trawling).

[0017] The feedback regulation stage adopts a closed loop of voltage sampling-formula calculation-signal output. The sampling resistor uses a high-precision alloy resistor (0.1% accuracy) to ensure that the voltage sampling error is ≤±0.1%. The surge protection stage is connected in series with a 10Ω current-limiting resistor to prevent surge current from damaging subsequent devices.

[0018] The high-voltage compensation formula is the core of this unit, and the formula is as follows: U_out=U_in×[1+k1×(P-0.1)+k2×|T-25|], (Formula 1); Dynamically offsets the negative impact of deep-sea temperature and pressure variations on the output voltage of the isolated power supply. Under 100MPa high voltage, the dielectric constant of the internal capacitor in a traditional power supply decreases by about 20%, and the output voltage ripple spikes from 20mV to 80mV; For every 10°C deviation of temperature from 25°C, the resistance value changes by about 3%, further exacerbating voltage fluctuations.

[0019] This formula calculates the voltage adjustment by real-time acquisition of P (pressure) and T (temperature) to stabilize the U_out ripple within 40mV.

[0020] Analysis and value basis of each parameter: Parameter symbol Parameter name unit Range of values Basis for value selection Physical meaning U_out Output voltage VDC 4.98-5.02V Based on the rated power supply requirements of each unit in the module, experiments verified that the optimal performance is achieved at 5V±0.02V. Provide stable power supply for subsequent units U_in Input voltage VDC 24-48V Submarine cable transmission loss test: 24V voltage drop ≤3V after 10km of cable, meeting input requirements. Input voltage of shore-based power supply system <![CDATA[k1]]> Pressure compensation coefficient 1 / MPa 0.0015-0.0025 <![CDATA[Autoclave experiment: within the range of 0.1 - 100 MPa, for every 1 MPa increase, U_out decreases by 0.002%. When fitting, the compensation error is minimized (±0.01 V) when k1 = 0.002]]> Counteracting the change in dielectric constant of capacitance caused by pressure P Real-time pressure MPa 0.1-100MPa The Keller 33X sensor's measurement range covers deep-sea applications from 0 to 10,000 meters. Real-time pressure values ​​in deep-sea environment 0.1MPa Reference pressure MPa Fixed value Standard atmospheric pressure, the reference environmental pressure for device calibration. Zero point of pressure change reference <![CDATA[k2]]> Temperature compensation coefficient 1 / ℃ 0.0008-0.0012 <![CDATA[High and low temperature chamber experiment: within the range of -40 to 80 °C, for every 1 °C deviation from 25 °C, U_out changes by 0.001%. When fitting, the compensation error is minimized (±0.005 V) with k2 = 0.001]]> To counteract temperature-induced changes in resistance T Real-time temperature ℃ -40-80℃ The AD590 sensor's measurement range covers the temperature range from the seabed surface to hydrothermal vents. Real-time temperature values ​​of deep-sea environment 25℃ Reference temperature ℃ Fixed value The standard calibration temperature of electronic devices is when their parameters are most stable. Zero point of temperature change The formula is derived based on a bivariate linear fitting of pressure-voltage and temperature-voltage. The experimental procedure is as follows: Pressure effect experiment: The module was placed in a HAST-100 autoclave at a fixed temperature of 25℃. The pressure was gradually increased from 0.1MPa to 100MPa, and the U_out value was collected every 10MPa. The data is as follows (excerpt): Pressure P (MPa) 0.120 40 60 80 100 U_out (V) 5.00 4.98 4.96 4.94 4.92 4.90 The fitted relationship between pressure and U_out is as follows: U_out = 5.00 - 0.001 × (P - 0.1), meaning that for every 1 MPa increase in pressure, U_out decreases by 0.001V, requiring a compensation of +0.001V / MPa, corresponding to k1 = 0.002 (because when U_in = 24V, 24 × 0.002 × (P - 0.1) ≈ 0.001 × (P - 0.1)).

[0021] Temperature effect experiment: The module was placed in a high and low temperature chamber with a fixed pressure of 0.1 MPa. The temperature was increased from -40℃ to 80℃, and the U_out value was collected every 20℃. The data is as follows (excerpt): Temperature T (°C) -40 -20 25 60 80 U_out (V) 4.97 4.98 5.00 4.99 4.98 The fitted relationship between temperature and U_out is: U_out = 5.00 - 0.0005 × |T-25|, which needs to be compensated by +0.0005V / °C, corresponding to k2 = 0.001 (24 × 0.001 × |T-25| ≈ 0.0005 × |T-25|).

[0022] Temperature-pressure coupling verification: Under extreme conditions of P=100MPa and T=-40℃, without compensation, U_out=4.85V, calculated using Formula 1 (U_in=24V, k1=0.002, k2=0.001): U_out=24×[1+0.002×(100-0.1)+0.001×|-40-25|]=24×[1+0.1998+0.065]=24×1.2648≈30.355V; In the formula, U_in is the input voltage, and the output voltage needs to be adjusted to 5V via feedback. The adjustment amount of U_out is: ΔU = U_in × [k1 × (P - 0.1) + k2 × |T - 25|], which is superimposed on the reference output of 5V, i.e., U_out = 5 + ΔU, to ensure that the output is stable at 5V ± 0.02V.

[0023] Surge protection subunit: A synergistic solution using MOV, GDT, and current-limiting resistors is adopted: When the surge voltage is ≤560V, the MOV varistor breaks down and discharges current. When the surge voltage is greater than 560V, the GDT gas discharge tube breaks down, further discharging the large current. A 10Ω current-limiting resistor in series prevents the leakage current from exceeding the MOV's rated value of 15kA. Experimental verification: After applying an 8 / 20μs, 15kA surge current, the sub-unit output voltage fluctuation was ≤0.5V, and it recovered to 5V within 100μs, without causing damage to subsequent units.

[0024] Photoelectric conversion core unit: This unit is the core of electro-optical and optical-electrical conversion. Its function is to convert electrical signals into 1550nm optical signals (the wavelength with the least loss) at the transmitting end and to convert the optical signals back into electrical signals at the receiving end. The core is the temperature-pressure coordinated bias current calibration formula, which solves the problem of optical power fluctuation caused by traditional fixed bias current.

[0025] Circuit architecture design: The unit is divided into a transmitting branch and a receiving branch: Transmitter branch: The CreeCGD15000GaN driver chip receives the bias current signal from the control unit, drives the JDSU1550nmLD laser diode to emit light signals, and monitors the current with a 10Ω resistor in series. Receiver branch: The HAMAMATSU G8371-01 InGaAsPIN photodiode converts the optical signal into a weak current, which is then amplified into a voltage signal by the ADIAD8011 preamplifier, and then filtered to remove noise.

[0026] The transmitting branch adopts a constant current drive mode to avoid optical power instability caused by current fluctuations in the LD; The receiving branch uses a low-noise preamplifier with an input noise voltage ≤1nV / √Hz to ensure effective amplification of weak optical signals.

[0027] The bias current calibration formula is as follows: I_b = I_b0 × [1 + 0.0003 × (P - 0.1) - 0.0002 × (T - 25)], (Formula 2).

[0028] The optical power output of an LD laser diode is directly related to its bias current (I_b), and the temperature and pressure environment of the deep sea can cause changes in the LD threshold current. High voltage deforms the LD junction region, causing the threshold current to rise; High temperatures increase the carrier concentration in semiconductors, causing the threshold current to decrease.

[0029] The traditional fixed I_b scheme exhibits an optical power fluctuation of ±0.8 dBm at P=100 MPa and T=-40℃. This formula uses temperature and pressure to calibrate I_b, controlling the fluctuation within ±0.05dBm.

[0030] Analysis and value basis of each parameter: Parameter symbol Parameter name unit Range of values Basis for value selection (experimental data) Physical meaning I_b Post-calibration bias current mA 48-52 JDSULD Experiment: At I_b=50mA, the optical power is 12dBm (rated value), and the optical power fluctuation is ±0.05dBm within ±2mA. Drive LD to emit stable optical power <![CDATA[I_b0]]> Reference bias current mA 50 Current value of LD output at 12dBm optical power at 25℃ / 0.1MPa Bias current reference 0.0003 / MPa Pressure calibration coefficient 1 / MPa Fixed value High voltage experiment: For every 1 MPa increase, the LD threshold current rises by 0.03%, requiring an additional 0.03% I_b to offset it, i.e., 0.0003 × (P - 0.1). Compensation for the rise in LD threshold current caused by high voltage -0.0002 / ℃ Temperature calibration coefficient 1 / ℃ Fixed value High and low temperature experiments: For every 1°C increase, the LD threshold current decreases by 0.02%, which requires a 0.02% reduction in I_b to offset, i.e., -0.0002×(T-25). Compensation for temperature-induced drop in LD threshold current P, T Real-time pressure and temperature MPa, ℃ Same as formula 1 Same as Formula 1, collected by Keller33X+AD590 Environmental parameter input Experiment on the effect of pressure: With the temperature fixed at 25℃, the pressure was increased from 0.1MPa to 100MPa, and the change in the LD threshold current (I_th) was measured: P (MPa) 0.120406080100 I_th (mA) 2020.1220.2420.3620.4820.60 Fitting yields I_th = 20 + 0.006 × (P - 0.1). To maintain stable optical power, I_b needs to increase synchronously with I_th, i.e., the pressure adjustment coefficient of I_b is 0.006 / 20 = 0.0003 / MPa (since I_b0 = 50mA, I_th = 20mA, the ratio is 2.5:1).

[0031] Temperature effect experiment: With the pressure fixed at 0.1 MPa, the temperature is increased from -40℃ to 80℃, and the change in I_th is measured: T (°C) -40 -20 25 60 80 I_th (mA) 20.48 20.24 20.00 19.84 19.76 Fitting the result, I_th = 20 - 0.004 × (T - 25), therefore the temperature regulation coefficient of I_b is -0.004 / 20 = -0.0002 / °C.

[0032] Temperature-pressure coupling verification: At P = 100 MPa and T = -40℃, I_b0 = 50 mA, substituting into formula 2: I_b=50×[1+0.0003×(100-0.1)-0.0002×(-40-25)]=50×[1+0.02997+0.013]=50×1.04297≈52.15mA.

[0033] Experimental verification: At this time, the LD output optical power is 12.02dBm with a fluctuation of only +0.02dBm, which is far better than the uncompensated 11.2dBm (fluctuation of -0.8dBm).

[0034] Signal optimization unit: This unit is a crucial component for improving signal quality. Its function is to equalize and suppress noise in the electrical signal reconstructed at the receiver, thereby reducing distortion. The core innovation is a self-developed distortion hierarchical optimization algorithm that solves the problems of inefficiency or undercompensation in traditional fixed-order equalization.

[0035] The unit adopts a four-stage architecture: distortion detection, equalization processing, noise suppression, and signal amplification. Distortion detection: The STM32H743VIT6 main control chip acquires input and output signals through ADC and calculates total harmonic distortion (THD). Equalization processing: ADIAD9250 adaptive equalizer, supporting 8th-order / 16th-order equalization mode switching; Noise suppression: ADIAD8421 differential amplifier (common-mode rejection ratio 85dB) + 50 / 60Hz notch filter; Signal amplification: TIOPA847 operational amplifier, adjustable gain from 0-30dB.

[0036] Analysis and Implementation of Distortion Grading Optimization Algorithm: Core algorithm logic: Based on the signal distortion (THD) classification, an equalization strategy is dynamically selected to achieve a balance between low distortion and high efficiency, and high distortion and high accuracy. The algorithm flow is as follows: Signal input → THD detection → THD ≤ 0.3%? → Yes: 8th order equalization; No → THD≤0.6%? → Yes: 16th order + predistortion; No: Redundancy switching → Noise suppression → Output.

[0037] Algorithm hierarchical logic and parameter settings: Distortion level THD range Optimization strategy Core parameters Processing delay Optimization effect Level 1 (Low Distortion) THD≤0.3% 8th order equilibrium Convergence speed 10μs, balanced bandwidth 1-50GHz ≤10ns THD reduced to ≤0.2%, SNR ≥58dB Level 2 (Low Distortion) 0.3% < THD ≤ 0.6% 16th-order equalization + predistortion correction The predistortion coefficients are calculated in real time based on the LMS algorithm, with a step size of 0.01. ≤20ns THD reduced to ≤0.25%, SNR ≥58dB Level 3 (Low Distortion) THD > 0.6% Redundant channel switching The switching signal is output by GPIO, and the switching time is ≤5μs. ≤5μs After switching, THD ≤ 0.3%, SNR ≥ 58dB Implementation of key algorithm steps: The algorithm is implemented using the FPU floating-point unit of the STM32H743VIT6 and FPGA coprocessor. The core steps and code logic are as follows: Figure 1 As shown.

[0038] Experimental Verification: Comparison of Algorithm with Traditional Approaches In a scenario with a transmission rate of 40Gbps and a signal distortion THD of 0.5% (medium distortion), the algorithm is compared with the traditional fixed 16th-order equalization scheme: index Algorithm of this invention (secondary optimization) Traditional fixed 16th order equalization Increase Processing delay 15ns 25ns 40% THD (optimized) 0.22% 0.30% 26.7% SNR 59.2dB 56.8dB 4.2% CPU utilization 35% 60% 41.7% Temperature and pressure coordinated control unit: This unit is the brain of the module, its function being to collect environmental and operating parameters, run formulas 1 and 2 and the distortion classification algorithm, and achieve coordinated control of various units. Its core is multi-task real-time scheduling and closed-loop feedback control.

[0039] Hardware and software architecture: Hardware core: STM32H743VIT6 main control chip, equipped with 16MB Flash (program storage) and 4MB SRAM (data cache), peripherals include 16-bit ADC (sampling rate 1MSPS), DAC (output precision 12 bits), and SPI / I2C communication interface (connecting sensors and equalizers).

[0040] Software architecture: Based on the RT-Thread real-time operating system, it adopts an "interrupt + task" scheduling mode. The core tasks and their priorities are as follows: Task Name Priority Execution cycle Core Functions CPU resources used Data acquisition task Highest (1) 1ms Collect parameters such as P, T, U_out, I_b, and THD. 20% Formula calculation task Second highest (2) 10ms Run formulas 1 and 2 to calculate the adjustment amounts U_out and I_b. 15% Algorithm control task Medium (3) 10ms Run the distortion grading algorithm to control the equalizer mode. 25% Health monitoring task Minimum (4) 100ms Run the health model and output early warning signals. 10% Implementation of closed-loop control logic: Taking U_out adjustment as an example, the closed-loop control logic flow is as follows: The data acquisition task collects U_out (sampled by ADC, 16-bit precision) and P and T every 1ms; The formula calculation task substitutes into Formula 1 every 10ms to calculate the target value of U_out; The adjustment amount ΔU = target value - actual value is calculated, and the analog voltage signal is output to the feedback regulation stage of the high voltage adapter unit through the DAC; The ADIOP07 operational amplifier in the feedback regulation stage converts ΔU into a current signal to control the feedback terminal of the isolated power supply; The isolated power supply adjusts the output voltage according to the feedback signal to stabilize U_out at the target value; The data acquisition task collects U_out again to verify the adjustment effect and form a closed loop.

[0041] Experimental verification: Under conditions of P=80MPa and T=60℃, an input voltage fluctuation of ±0.5V was artificially introduced. After the closed-loop control was started, U_out was adjusted from the initial fluctuation of 4.95V to 5.00V±0.01V, with an adjustment time of ≤20ms, which is far better than the fluctuation range of 4.80-5.10V without closed-loop control.

[0042] Anti-interference design for temperature and pressure coordinated control: The deep-sea environment is subject to electromagnetic interference (such as power frequency interference induced by submarine cables) and mechanical vibration (such as ocean current impact), requiring targeted anti-interference measures to be designed. Hardware anti-interference: Twisted pair cable is used to transmit sensor signals to reduce electromagnetic coupling; a common mode inductor (TDKACM2012-900-2P) is added at the power input to suppress common mode noise; Software anti-interference: Data acquisition uses a three-sample averaging algorithm to remove outliers; task scheduling uses a time-slice round-robin mechanism to avoid priority inversion; Mechanical anti-interference: The sensor is fixed by a metal bracket, and the space between the bracket and the housing is filled with damping rubber (Shore hardness 60 degrees).

[0043] Anti-interference test: Under electromagnetic interference environment of 1kHz and 1V / m, without anti-interference measures, the sampling errors of P and T are ±0.5MPa and ±2℃, respectively; after taking measures, the errors are reduced to ±0.2MPa and ±0.8℃, which meet the control accuracy requirements.

[0044] Multi-dimensional status monitoring unit: The multi-dimensional status monitoring unit is integrated into the temperature and pressure co-control unit. Its core function is to solve the problem of high misjudgment rate in traditional single-parameter monitoring by integrating the health status of multi-parameter assessment modules. Its innovation lies in the creation of a self-developed health assessment model (Formula 3), which enables early warning and accurate diagnosis.

[0045] The formula for the health assessment model is as follows: H=0.35×(P_out / P_out0)+0.3×(S / N) / (S / N)0+0.2×(I_b / I_b0)+0.15×(U_out / U_out0), (Formula 3).

[0046] Formula 3 uses a weighted fusion of four core parameters—output optical power (P_out), signal-to-noise ratio (S / N), bias current (I_b), and output voltage (U_out)—to derive a health index H (better closer to 1). Traditional monitoring focuses only on U_out; fluctuations of ±0.1V in U_out may be misjudged as faults. This model, however, combines key performance parameters such as P_out and S / N, reducing the misjudgment rate from the traditional 15% to ≤2%.

[0047] Analysis of each parameter and the basis for its weight: Parameter symbol Parameter name unit Range of values Weighting coefficient Weighting basis (AHP) H Health - [0,1] - Based on comprehensive evaluation indicators, an H ≤ 0.85 triggers an early warning, and an H ≤ 0.75 triggers a fault. <![CDATA[P_out / P_out0]]> Normalized value of optical power - [0.8,1.2] 0.35 AHP has the greatest impact on transmission quality and is ranked first in terms of weight. <![CDATA[(S / N) / (S / N)0]]> Normalized signal-to-noise ratio - [0.8,1.2] 0.3 Affects signal anti-interference ability, second in weight ranking <![CDATA[I_b / I_b0]]> Normalized bias current - [0.9,1.1] 0.2 Reflecting the working status of the LD, ranked third in weight. <![CDATA[U_out / U_out0]]> Normalized output voltage - [0.95,1.05] 0.15 Power supply stability index, ranked fourth by weight. Formula derivation and experimental verification: Derivation of weighting coefficients (AHP method): Establish a hierarchical structure: Target layer (health level H) → Criteria layer (4 parameters) → Solution layer (module status); Construct a judgment matrix: Determine the importance ratio between parameters by scoring them with scores from 10 experts in the field of communications (e.g., P_out is more important than S / N, with a ratio of 1.2). Consistency test: The random consistency index CR is calculated to be 0.04 < 0.1, indicating that the weights are effective. The final weight coefficients are 0.35, 0.3, 0.2, and 0.15.

[0048] Normalization processing logic: Taking P_out as an example, P_out0 = 12dBm (rated value). When P_out = 10.8dBm, the normalized value = 10.8 / 12 = 0.9. If P_out > 1.2 × P_out0 or < 0.8 × P_out0, it is directly judged as abnormal, and H is forcibly set to 0.7.

[0049] Experimental verification: Simulating three typical states to test the accuracy of the model: Simulation state Output optical power P_out (dBm) Signal-to-noise ratio (S / N) (dB) Bias current I_b (mA) Output voltage U_out (V) Health score H calculated value Actual state Judgment result normal 12.05 59.05 50.00 5.00 0.98 normal correct Minor abnormality (warning) 12.05 57.04 51.98 4.80 0.84 LD (laser diode) aging Correct early warning Serious anomaly (malfunction) 9.55 55.53 53.04 4.90 0.72 LD failure Correct alarm By simulating three states of the module—"normal, slight aging, and severe fault"—and inputting parameters such as output optical power, signal-to-noise ratio, bias current, and output voltage, the health status H is calculated using the health status formula. This verifies that the model can accurately identify different states (normal, warning, and fault), providing a reliable basis for subsequent early warning mechanisms and operation and maintenance decisions.

[0050] Verification conclusion: In 100 sets of simulation tests, the fault identification accuracy of the health status model of this invention was 98%, and the false judgment rate was 2%; the accuracy of traditional single-parameter monitoring was 75%, and the false judgment rate was 15%. The model effectively improved the accuracy of status assessment.

[0051] Early warning mechanism: The early warning mechanism based on the health assessment model adopts a layered architecture of local rapid response + remote collaborative scheduling to avoid the delay or false triggering problems of a single early warning mode. The specific design is as follows: Warning level classification and triggering logic: Given the high maintenance costs of deep-sea shipping, a two-tiered early warning threshold is set to balance sensitivity and fault tolerance: Level 1 Warning (Minor Abnormality): Health level 0.75 < H ≤ 0.85. The triggering condition is a single parameter deviating from the rated value by 5%-10% (e.g., P_out = 10.8-11.4dBm), or the health level decreasing by ≥0.05 for 3 consecutive cycles (300ms).

[0052] Level 2 warning (serious anomaly): Health H ≤ 0.75. The trigger condition is that any parameter deviates from the rated value by more than 10%, or two or more parameters deviate from the rated value by 5%-10% at the same time (such as P_out < 10.8dBm and S / N < 57dB).

[0053] Early warning execution and data reporting process: Local warning execution: The module integrates a red LED indicator and a low-power buzzer (suitable for shallow seas where sound can be heard), driven by the GPIO port of the STM32H743VIT6 main control chip. Level 1 warning: LEDs flash at a frequency of 1Hz, and buzzers are silenced (to avoid interfering with surrounding observation equipment); Level 2 warning: LED is constantly lit, buzzer sounds intermittently at a frequency of 2Hz, power consumption ≤50mW.

[0054] Remote data reporting: Warning information is packaged in the format of "timestamp (accurate to milliseconds) + parameter snapshot + H value + warning level", synchronized to the photoelectric conversion module via the SPI interface, and uploaded to the shore-based main control via fiber optic link. The reporting cycle for normal status is 1 second. When a level 2 warning is triggered, an interrupt reporting is performed with a delay of ≤0.3 seconds. Data integrity is ensured by using CRC-16 checksum.

[0055] Historical data traceability: The main control chip is equipped with 16MB Flash storage for nearly 1,000 historical status data (including 30 snapshots before and after the warning), which can be read through AT commands issued from the shore base, providing data support for fault tracing, and the data retention time is ≥30 days.

[0056] Data transmission system: Based on a high-voltage photoelectric conversion module, a three-level submarine optical communication system consisting of module nodes, intelligent relays, and shore-based main control is constructed to achieve high-speed transmission of 10-60Gbps, adapting to high voltage of 30-100MPa and temperature variation environments of -40-80℃. The core solution addresses the problems of signal attenuation, dispersion distortion, and bandwidth allocation in long-distance transmission.

[0057] System overall architecture and network design: A hybrid topology of star branch + chain backbone is adopted. The functions of each unit and the networking parameters are as follows: Module nodes: 32 transceiver module pairs, which are connected to deep-sea observation equipment (seismographs, high-definition cameras, etc.) to realize electro-optical / optical-electrical conversion and synchronously upload health status data; Intelligent relay unit: One unit is deployed every 100km. Its core functions are optical power amplification (adjustable gain from 0-40dB) and dynamic dispersion compensation to solve long-distance transmission loss. shore-based main control unit: Deployed at coastal base stations, it enables dynamic bandwidth allocation, full system status monitoring, and AI-based fault prediction; Transmission link: It adopts armored single-mode optical cable (Corning SMF-28e+), with a wavelength attenuation of ≤0.18dB / km at 1550nm and a tensile strength of ≥10kN for the outer steel wire armor.

[0058] The core components have undergone high-voltage compatibility testing, and the key parameters are shown in the table below: unit Key component models Core parameters (verified at 100MPa high pressure) Deployment quantity Module Node Custom development (detailed in Chapter 4) Optical power 12±0.05dBm, data rate 10-60Gbps 64 units (32 pairs) Smart Relay Huawei OptiXOSN1800V (Customized) Dispersion compensation 500-2000ps / nm, 16 ports 2 units (100km link) shore-based main control Xilinx XC7K325T + Jetson TX2 Processing latency ≤3μs, cache 2GB 1 unit Intelligent relay unit: The intelligent relay unit is the core of link compensation. Traditional fixed dispersion compensation schemes deteriorate the bit error rate to 10⁻¹ at a rate of 60Gbps. 0 This enables coordinated adaptation of temperature, pressure, speed, and distance.

[0059] Circuit architecture and control flow: It adopts a closed-loop architecture of light monitoring, dispersion analysis, compensation adjustment, and power amplification, with the core controlled by an FPGA. Optical power monitoring: The Agilent 81635A optical power meter collects the input optical power with a sampling accuracy of ±0.01dBm and a period of 1ms; Dispersion analysis: The FPGA's built-in FFT module analyzes the signal spectrum and calculates the pulse broadening (a core dispersion indicator). Compensation adjustment: The target compensation amount is calculated based on Formula 4, and the fiber optic grating compensator is driven by the ADIAD9910DDS chip; Power amplification: The JDSUEDFA optical amplifier dynamically adjusts the gain to ensure that the output power is stable at 13±0.1dBm.

[0060] Dispersion dynamic compensation formula: D = D0 × [1 + k_L × (L - 50) + k_R × (R - 30)], (Formula 4) Parameter analysis and value selection criteria: Parameter symbol Parameter name unit Range of values Basis for value selection (supported by experimental data) D Dynamic dispersion compensation ps / nm 500-2000 The fiber Bragg grating compensator has a hardware range covering 10-100km / 10-60Gbps across all scenarios. <![CDATA[D0]]> Baseline compensation amount ps / nm 1000 <![CDATA[Optimal compensation amount when L = 50 km and R = 30 Gbps. Experimental verification shows that the bit error rate is the lowest at this time (5×10⁻¹ 5 ).]]> k_L Distance compensation coefficient 1 / km 0.002 High-pressure environment experiment: For every 1km increase in L, the dispersion increases by 0.2%, requiring a corresponding increase in compensation of 0.2%. L Transmission distance km 10-100 The actual length of the optical cable during deployment is transmitted from the shore-based main control unit to the relay. k_R Rate compensation coefficient 1 / Gbps 0.001 Rate experiment: For every 1Gbps increase in R, the impact of dispersion on the signal increases by 0.1%, requiring a 0.1% increase in compensation. R Transmission rate Gbps 10-60 The photoelectric conversion module reports to the relay unit, supporting dynamic adjustment. Formula derivation and experimental verification: Coefficient derivation: In the controlled variable method experiment, with R=30Gbps fixed, the dispersion was fitted to obtain 20×L (ps / nm). k_L=(20×1 / 20×50) / 1=0.002 / km; With L = 50 km fixed, the fitted influence coefficient is 0.02 × R, and k_R = 0.001 / Gbps.

[0061] Extreme environment verification: When P=100MPa, T=-40℃, L=100km, R=60Gbps; D=1000×[1+0.002×50+0.001×30]=1130ps / nm.

[0062] Experimental results: Pulse widening 12 ps, bit error rate 8 × 10⁻¹ 5 ; Traditional fixed-compensation (D=1000ps / nm) pulse broadening is 35ps, with a bit error rate of 2×10⁻¹. 0 This solution improves the bit error rate by 5 orders of magnitude.

[0063] shore-based main control unit: The shore-based master control is the decision-making center of the system, solving the problem that traditional evenly distributed bandwidth cannot meet the needs of high-priority services (such as real-time video).

[0064] Hardware and software architecture: Hardware: FPGA+GPU heterogeneous architecture, with the FPGA responsible for 60Gbps signal processing (latency ≤3μs) and the GPU running the LSTMAI model to predict faults 72 hours in advance; Software: Linux 4.14 kernel, multi-threaded scheduling (data reception > bandwidth allocation > status monitoring > AI prediction), inter-thread shared memory communication, latency ≤1μs.

[0065] Bandwidth dynamic allocation formula: With a total system bandwidth of B_total = 60Gbps and 32 nodes accessing the network in parallel, the differentiated allocation formula is as follows: B_i=B_total×(α_i×Q_i) / Σ(α_i×Q_i), (Formula 5); Parameter parsing and application examples: Parameter symbol Parameter name Value selection rules and basis B_i The i-th node is allocated bandwidth. ≥1Gbps (minimum guaranteed bandwidth), ≤20Gbps (maximum bandwidth per node) B_total Total system bandwidth Fixed at 60Gbps (determined by the hardware bandwidth of the intelligent repeater unit). α_i Business priority coefficient Real-time video transmission α=1.0, data acquisition α=0.5, status reporting α=0.3 (set based on business latency requirements) Q_i Data volume coefficient Q_i = Real-time data volume of the node / 1Gbps (e.g., 10Gbps of data corresponds to Q=10, which is reported by the node in real time) Σ(α_i×Q_i) Weighted sum The sum of α_i × Q_i of all access nodes is used as a normalization factor. Application example: The weighted sum of the bandwidth allocated to the three typical nodes—Node 1 (real-time video, α=1.0, Q=10), Node 2 (data acquisition, α=0.5, Q=5), and Node 3 (status reporting, α=0.3, Q=1)—is 1.0 × 10 + 0.5 × 5 + 0.3 × 1 = 12.8 GHz. B1 = 60 × 10 / 12.8 ≈ 46.875 Gbps (meets video requirements), B2 ≈ 11.719 Gbps, B3 ≈ 1.406 Gbps (higher than the minimum guarantee).

[0066] System-level experimental verification and reliability testing: A 100km seabed environment simulation platform (autoclave + high and low temperature chamber + electromagnetic interference generator) was constructed, and three sets of implementation examples and two sets of comparative examples were designed to verify its performance. Example 1: Shallow sea normal environment (P=30MPa, T=25℃, R=30Gbps) Configuration: 32 nodes (10 video, 22 data acquisition), continuously for 24 hours; Results: Module U_out = 5.00 ± 0.01V, relay bit error rate 5 × 10⁻¹ 5 The video node bandwidth is 5.2-5.8Gbps, with no packet loss and 100% stability.

[0067] Example 2: Deep-sea extreme environment (P=100MPa, T=-40℃, R=60Gbps) Configuration: 16 high-priority video nodes, continuously for 12 hours; Results: After compensation using formulas 1 and 2, the module's U_out = 4.99 ± 0.02V, relay D = 1130ps / nm, and bit error rate 8 × 10⁻¹ 5 The node bandwidth is 3.75Gbps, and the packet loss rate is 0.05%.

[0068] Example 3: Fault simulation environment (P=80MPa, T=60℃, node 5LD aging) Configuration: 32 nodes, with the optical power of node 5 manually reduced to 10.8 dBm; Results: Node 5H=0.84 triggered a Level 1 warning, with a shore-based delay of 0.3s; after automatically adjusting the bandwidth to 4Gbps, the bit error rate decreased from 10⁻¹² to 10⁻¹. 4 The AI ​​model can predict aging 24 hours in advance.

[0069] Comparative Example 1: Traditional fixed parameter scheme (without formulas 1, 2, and 4) Conditions: Same as in Example 2; Results: Module U_out = 4.85 ± 0.05 V, relay bit error rate 2 × 10⁻¹ 0 Data packet loss rate of 3%, video stuttering.

[0070] Comparative Example 2: Traditional single-parameter monitoring (monitoring only U_out) Conditions: Same as in Example 3; Results: Node 5U_out=4.98V (normal), no warning; the bit error rate rose to 10⁻¹¹, causing congestion on 3 nodes, and manual investigation took 2 hours.

[0071] System reliability and environmental adaptability design: Redundancy backup: Module 1 primary and 1 backup (5μs switching), relay dual power supply (lithium battery life 72h), shore-based dual-machine hot standby (switching delay ≤10ms); Environmental adaptability: The core results of 10 extreme tests are shown in the table below: Test Project condition result High voltage test 100MPa for 1000h Performance degradation ≤1%, no leakage High and low temperature cycling Cycle 100 times between -40℃ and 80℃. Parameter fluctuation ≤ ±1% Salt spray test 5% salt spray for 500 hours Non-corrosive Vibration test 10-2000Hz, 10g No performance fluctuation Lifespan Guarantee: Accelerated aging tests estimate a lifespan of ≥10 years; key components are selected using long-life models (LD lifespan ≥10 years). 5 Hour).

[0072] Temperature and voltage coupling dynamic compensation: Formulas 1 and 2 offset environmental influences, voltage ripple ≤40mV, optical power fluctuation ≤±0.05dBm, and stability improved by 50%; Distortion grading optimization: THD grading adaptation and balancing strategy, processing latency ≤20ns, SNR≥58dB, CPU utilization reduced by 41.7%; Multi-dimensional health model: Formula 3 integrates 4 parameters, with a fault identification accuracy of ≥98% and a false positive rate of ≤2%, which is 120% higher than the traditional model; End-to-end adaptive adjustment: Formulas 4 and 5 optimize dispersion and bandwidth, achieving a bit error rate of ≤10⁻¹ at 60Gbps. 4 This represents an improvement of 4-5 orders of magnitude compared to traditional methods.

[0073] Compared with existing technologies: index This invention Existing technology Advantages Pressure rating 30-100MPa ≤50MPa Covering the entire deep sea (0-10000 meters) Transmission rate 10-60Gbps ≤40Gbps Meets the requirements of high-definition video and large data transmission Bit error rate <![CDATA[≤10⁻¹ 4 ]]> <![CDATA[≥10⁻¹ 0 ]]> Stability improved by 4 orders of magnitude Fault identification accuracy ≥98% ≤75% Maintenance costs reduced by more than 60% This invention is applied to fields such as deep-sea oil and gas exploration and seabed observation networks, enhancing my country's international competitiveness in deep-sea communication technology. It is a high-voltage photoelectric conversion module and seabed optical communication data transmission system designed for extreme deep-sea environments. Specifically designed for high-voltage scenarios of 30-100MPa and temperature variations of -40-80℃, it achieves high-speed transmission of 10-60Gbps (bit error rate ≤10%) through core technologies such as dynamic compensation for temperature-pressure coupling, distortion classification optimization, multi-dimensional health assessment, and link adaptive adjustment. -14 With a fault identification accuracy of ≥98%, it solves the pain points of traditional solutions such as performance drift, high bit error rate, and difficult operation and maintenance. It can be widely used in deep-sea oil and gas exploration, seabed observation network and other fields, and has both technological leadership and significant economic and social benefits.

[0074] The embodiments of the present invention are given for illustrative and descriptive purposes only, and are not intended to be exhaustive or to limit the invention to the forms disclosed. Many modifications and variations will be apparent to those skilled in the art. The embodiments were chosen and described in order to better illustrate the principles and practical application of the invention, and to enable those skilled in the art to understand the invention and to design various embodiments with various modifications suitable for a particular purpose.

Claims

1. A high-voltage resistant photoelectric conversion module suitable for submarine optical communication, characterized in that: It includes a high-voltage adapter unit, a photoelectric conversion core unit, a signal optimization unit, and a temperature and pressure coordinated control unit; The high-voltage adapter unit is compatible with seabed pressure environments of 30-100MPa, with an input voltage of 24-48VDC, and the output voltage is adjusted by a high-voltage compensation formula. U_out=U_in×[1+k1×(P-0.1)+k2×|T-25|], where k1 is the pressure coefficient, k2 is the temperature coefficient, P is the real-time pressure, T is the real-time temperature, and the output ripple is ≤40mV; The photoelectric conversion core unit adopts an InGaAsPIN and GaN driver chip architecture, with a responsivity ≥0.9A / W at a wavelength of 1550nm, dark current ≤0.7nA, and bias current optimized by a calibration formula. I_b = I_b0 × [1 + 0.0003 × (P - 0.1) - 0.0002 × (T - 25)], where I_b0 is the reference bias current; The signal optimization unit has a balanced bandwidth of 1-50GHz, and the control unit has a response time of ≤35μs.

2. The module according to claim 1, characterized in that, The signal optimization unit employs a distortion hierarchical optimization algorithm: When the signal distortion is ≤0.3%, 8th-order equalization is used; When the distortion is between 0.3% and 0.6%, the system automatically upgrades to 16th-order equalization and initiates pre-distortion correction, resulting in a distortion of ≤0.25% after correction. When the distortion is greater than 0.6%, redundant channel switching is triggered, the switching delay is less than 5μs, and the signal-to-noise ratio of the processed signal is greater than 58dB.

3. The module according to claim 1, characterized in that, The high-voltage adapter unit includes an isolated power supply and a surge protection subunit; The isolation voltage of the isolated power supply is ≥5.5kVrms, and the power conversion efficiency meets the efficiency formula: η=0.92-0.0005×(P-0.1)-0.0003×|T-25|, η≥85%; The surge protection subunit can withstand 8 / 20μs surge current ≥15kA, has a response time ≤6ns, and recovers rated output within 100μs after the surge.

4. The module according to claim 1, characterized in that, It also includes a multi-dimensional status monitoring unit, which uses a health assessment model: H = 0.35 × (P_out / P_out0) + 0.3 × (S / N) / (S / N)0 + 0.2 × (I_b / I_b0) + 0.15 × (U_out / U_out0), H ∈ [0,1]. When H ≤ 0.85, an early warning is triggered; when H ≤ 0.75, a fault report is triggered. The monitoring error is ≤ ±1.5%.

5. A data transmission system suitable for submarine optical communication, characterized in that, Includes at least two of the high-voltage photoelectric conversion modules, intelligent relay units, shore-based main control units, and submarine optical transmission links as described in any one of claims 1-4; The module is connected to the intelligent relay unit via armored optical cable, with a transmission rate of 10-60Gbps. The intelligent repeater unit adopts a self-developed link loss compensation formula: L_comp = L_meas × [1 + 0.0006 × (T - 25) + 0.0004 × (P - 0.1)], where L_meas is the measured loss and the compensation accuracy is ≤ ±0.25dB; The shore-based master control unit achieves global parameter coordinated configuration, and the system steady-state bit error rate is ≤10. -14 .

6. The system according to claim 5, characterized in that, The intelligent relay unit includes an adaptive dispersion compensation subunit, and the compensation amount is calculated according to the formula: C_disp=D×L×R×0.93×[1+0.0002×(T-25)], where D is the fiber dispersion coefficient, L is the repeater length, R is the transmission rate, 0.93 is the environmental correction factor, the compensation accuracy is ≤±0.3ps / nm, and the repeater gain is 35-45dB.

7. The system according to claim 5, characterized in that, The shore-based main control unit includes a heterogeneous data processing subunit, an intelligent power supply management subunit, and a fault location subunit. The heterogeneous processing subunit adopts an FPGA+GPU architecture, with a data cache capacity of ≥1.5GB and a processing latency of ≤6μs; The power supply management subunit outputs 3500-6000VDC, with ripple ≤70mV and power supply efficiency ≥92%; The fault location subunit combines the module H value with OTDR data, achieving a location accuracy of ≤20m and a diagnostic accuracy of ≥98%.

8. The module according to claim 1, characterized in that, The module calibration method steps are as follows: 1) Place it in an analog environment and input a 1GHz, 1Vpp standard electrical signal; 2) Collect the output optical power P1 and dark current I_d; 3) Adjust the high voltage adapter output according to the calibration formula: ΔU=0.03×(I_d-0.5)×(P / 50), where ΔU is the voltage adjustment amount and 0.5 is the reference dark current; 4) Repeat steps 2-3 until P1 fluctuation is ≤ ±0.05dBm, then the calibration is complete.

9. The system according to claim 5, characterized in that, The data transmission steps are as follows: 1) The shore-based main control unit sends out initialization parameters, setting module I_b to I_b0 and relay gain to 40dB; 2) The transmitting module converts the electrical signal into a 10-15dBm optical signal and injects it into the transmission link; 3) The intelligent relay unit collects L_meas and T, P in real time, and amplifies them after compensation according to the L_comp formula; 4) The receiving module restores the electrical signal and processes it through the signal optimization unit; 5) BER monitoring by shore-based units; when BER > 10 -14 At that time, according to ΔG=0.4×(BER / 10 -14 Adjust the relay gain; response time ≤ 12μs.

10. The system according to claim 5, characterized in that, It also includes an intelligent redundancy unit, which uses a switching decision formula: S=0.4×(ΔP / ΔP0)+0.3×(BER / 10 -14 )+0.2×(1-H)+0.1×(ΔT / ΔT0), where ΔP is the optical power fluctuation and ΔT is the temperature fluctuation. When S≥0.65, the system automatically switches to the backup module / link with a switching time ≤6ms and an interruption time ≤30μs.

Images