Bias-controlled two-dimensional self-adaptive PID system and method for electro-optical modulator

Abstract

The invention discloses a bidimensional self-adaptive PID system and method for electro-optical modulator bias control. The bidimensional self-adaptive PID system comprises a signal acquisition module, a bidimensional identification module, a self-adaptive threshold value module, a PID adjustment module and a closed-loop optimization module. The signal acquisition module acquires modulator output feedback, optical power fluctuation and environmental interference signals; the two-dimensional recognition module recognizes the type of a working point, and error classification is completed in combination with the PID error change rate after filtering; the self-adaptive threshold module dynamically adjusts an error judgment threshold according to the type of the working point and the environmental interference intensity; the PID adjusting module performs linkage adjustment based on working point types, error characteristics and a half-wave voltage three-dimensional parameter mapping table, and performs stable control process through parameter smooth transition, error pre-correction and a double-correlation overshoot suppression algorithm during working point switching. And the closed-loop optimization module iteratively optimizes PID parameters through performance index feedback. According to the invention, the problems of poor parameter adaptability, weak interference resistance and switching overshoot pain point of the traditional PID are solved, and high-precision stable control of the working point is realized.

Description

Technical Field

[0001] This invention belongs to the field of optoelectronic control technology, specifically relating to a two-dimensional adaptive PID system and method for bias control of an electro-optic modulator. Background Technology

[0002] Electro-optic modulators are core components of optical communication and fiber optic sensing systems, responsible for loading electrical signals onto optical signals to achieve modulation. Their operating point stability directly determines modulation accuracy and system reliability. However, their transmission characteristics are cosine-like, making them susceptible to drift due to external interference such as temperature, vibration, and fluctuations in light source power, requiring real-time adjustment via bias control algorithms. Traditional fixed-parameter PID bias control suffers from drawbacks such as parameters not adapting to multiple operating points, lack of error rate filtering leading to misjudgment of noise, weak anti-interference capabilities due to linear threshold adjustment, insufficient integral accuracy, high risk of overshoot during operating point switching, and hysteresis in closed-loop parameter updates. Related patents also have shortcomings: Patent application number 202211340430.9 uses FFT and PID algorithms but does not dynamically adjust PID parameters, resulting in insufficient anti-interference and switching stability; while patent application number 201210562354.6 can identify the operating point position but does not address error grading filtering, nonlinear threshold adjustment, and parameter iterative optimization, leading to poor noise suppression. Existing technologies lack PID control schemes that can adapt to multiple operating points and different types of modulators, accurately distinguish between noise and real drift, and achieve smooth switching of operating points, making it difficult to meet the high-precision and stable control requirements in complex environments. To address this, a two-dimensional adaptive PID system and method for bias control of electro-optic modulators is proposed. Summary of the Invention

[0003] The purpose of this invention is to propose a two-dimensional adaptive PID system and method for bias control of electro-optic modulators, which solves the problems of poor parameter adaptability, weak anti-interference, and switching overshoot of traditional PID, and achieves high-precision and stable control of the operating point of different types of electro-optic modulators.

[0004] This invention is achieved through the following technical solution: A two-dimensional adaptive PID system and method for bias control of an electro-optic modulator is disclosed. The system includes a signal acquisition module, a two-dimensional recognition module, an adaptive threshold module, a PID adjustment module, and a closed-loop optimization module. The modules work together to achieve bias adaptive control.

[0005] Furthermore, the signal acquisition module includes a photoelectric detection module, a signal conditioning module, and an ADC module, used to continuously acquire the output feedback signal, input optical power fluctuation signal, and environmental interference signal of the electro-optic modulator. The photoelectric detection module uses a photodiode with a response speed of 1μs~50μs to convert the output optical signal of the electro-optic modulator into an electrical signal. The signal conditioning module includes an IV conversion amplifier circuit and a bandpass filter circuit to filter and amplify the electrical signal and remove noise interference. The ADC module uses a 16-bit AD7893 chip to convert the analog signal into a digital signal at a sampling rate of 102.4kSPS. The output feedback signal is a DC component signal or a frequency component power signal. The environmental interference signal includes temperature fluctuation signal and electromagnetic interference intensity signal, and the half-wave voltage of the electro-optic modulator is acquired simultaneously. It is pre-calibrated or detected in real time through a dedicated detection circuit to provide data support for subsequent modules.

[0006] Furthermore, the dual-dimensional recognition module implements two main functions: operating point type recognition and error classification recognition. Operating point type recognition calculates the slope of the electro-optic modulator's transmission characteristic curve based on the DC component signal D in the output feedback signal. The DC component signal D is given by the formula Calculation, where The electro-optic conversion coefficient, For input optical power, The half-wave voltage of the electro-optic modulator. Where A is the DC bias voltage, and A is the amplitude of the sinusoidal detection signal. Let t be the angular frequency of the sinusoidal detection signal, and t be time. The offset phase; the slope k and the DC component signal D satisfy a linear relationship. , The proportionality constant is determined through a three-step calibration method: First, a standard voltage signal with 0.1% accuracy is input at the linear operating point of the electro-optic modulator; second, the corresponding output feedback D value and the slope of the measured transmission characteristic curve are recorded. The third step is to calculate. Repeat the calibration 5 times and take the average value; the calibration error should not exceed 2%; the judgment rule is when... When it is determined to be a linear operating point, When it is determined to be an extreme operating point, This is the slope threshold, with a value range of 0.01~0.05W / V and a default value of 0.02W / V.

[0007] Furthermore, the error classification identification first calculates the PID error signal. , Set a value for the target. To provide feedback on the actual value, a first-order difference combined with a 3-point moving average filter is used to calculate the rate of change of error. The approximate formula is: The approximate formula includes the signal acquisition period. , The value range is 1~10ms, with a default value of 5ms; the grading judgment criterion is based on the dynamic threshold output by the adaptive threshold module, when... and When it is determined to be noise error, no adjustment is needed; when and When it is determined to be a slow drift error, it is adjusted gently; when and It is immediately identified as a sudden drift error and responds quickly.

[0008] Furthermore, the adaptive threshold module dynamically adjusts the error threshold according to the operating point type and the intensity of environmental interference to avoid misjudgment caused by a fixed threshold; the initial threshold is set differently according to the operating point type: initial amplitude threshold for linear operating points. Initial rate of change threshold Extreme operating point initial amplitude threshold Initial rate of change threshold Define the environmental disturbance intensity coefficient. ,in The amplitude of the input optical power fluctuation. The rated input optical power is 10mW. This represents the amplitude of temperature fluctuation. Rated operating temperature 25℃; non-linear adjustment coefficient , From the formula , The calculation shows that its changing pattern is as follows: When it is a weak interference, , To perform threshold fine-tuning; When it is strong interference, , The threshold is significantly increased to avoid false triggering due to interference; the threshold adjustment formula based on the intensity of environmental interference is as follows: , At the same time, set the threshold constraint as , ,in , , , To avoid over-adjustment that could lead to recognition failure.

[0009] Furthermore, the PID control module pre-stores operating point type, error characteristics, and half-wave voltage. A three-dimensional PID parameter mapping table is used to map the half-wave voltage based on the dual-dimensional identification results and the signal acquisition module. Precisely call the proportional coefficient Integral time constant The matching adjustment strategy is implemented, and when switching operating points, a parameter smoothing transition of the dynamic time constant, error pre-correction, and a dual correlation overshoot suppression algorithm for amplitude and rate of change are performed. The specific steps are as follows: Step S1: The grading is based on the setting of the half-wave voltage of the electro-optic modulator. Based on the core classification criteria, it is divided into: Small half-wave voltage, suitable for high-speed response modulators; Medium half-wave voltage, compatible with general-purpose modulators; Large half-wave voltage, compatible with high-power stable lock-in modulators, each range is configured with dedicated reference parameters and adjustment strategies; Step S2 The gear adjustment strategy, targeting the linear operating point, will not adjust if the error characteristic is noise interference; otherwise, it will use a slow drift characteristic. , Perform standard PI control to gently correct drift. If the error characteristic is abrupt drift, then use... , Execute enhanced response adjustment; for extreme operating points, noise error is not adjusted, while slow drift error is adjusted. , Sudden change drift error is adopted , The adjustment formula is unified as follows ; Step S3 Gear adjustment strategy: slow drift error of linear operating point , Sudden change drift error , Extreme operating point slow drift error , Sudden change drift error , , parameter adaptation modulator response characteristics for medium half-wave voltage; Step S4 Gear adjustment strategy: slow drift error of linear operating point , Sudden drift error , Extreme operating point slow drift error , Sudden change drift error , By reducing Increase Adapted to the slow response characteristics of large half-wave voltage modulators; Step S5, High-precision integral calculation: The integral calculation of all PI controls adopts Simpson's integral method, and the specific approximate formula is as follows: ,in Where N is the signal acquisition period and N is the number of integration steps, N≥10 can meet the accuracy requirements and effectively avoid the integration error of traditional trapezoidal integrals in scenarios with rapidly changing errors.

[0010] Furthermore, the PID control module employs a three-step strategy for switching its operating point: smooth parameter transition, error pre-correction, and overshoot suppression. The specific steps are as follows: Step S1: Parameter smooth transition setting. The switching trigger signal is set to be output from the interrupt pin of the MCU control module, and the switching time... Precisely synchronized to the start point of the ADC sampling period; current operating point parameters. , With target parameters , All data is stored in the MCU's on-chip Flash memory, and can be quickly read using dual indexes of operating point type and half-wave voltage; dynamic transition time for operating point switching. The parameter switching amplitude is determined by the transition time. The larger the transition amplitude, the longer the transition time and the better the parameter smoothness. Define the parameter switching amplitude. , Dynamic transition time The value ranges from 50 to 200 ms. During the transition, the parameter is updated according to an exponential smoothing law, and the change formula is as follows: , The parameter update frequency is consistent with the ADC sampling rate to ensure that the change curve is continuous without any breaks; when hour, , The parameters are directly switched to the target value, and a parameter verification interruption is triggered. Step S2: Error pre-correction calculation. The pre-sampling process is started 10ms before the switch, and the current bias voltage is calculated by the ADC module. After 16 consecutive data acquisitions and filtering using a moving average, an accuracy of ±0.005V was obtained. Value; Target operating point bias reference value The query is performed from the three-dimensional parameter mapping table, which uses a two-dimensional array storage structure. The first dimension is the operating point type: linear / extreme, and the second dimension is the half-wave voltage level. Initial error correction amount Corrected initial error The correction coefficient of 0.8 was determined through multiple sets of comparative experiments: when the coefficient is 0.7, insufficient pre-correction leads to a slow response after switching; when it is 0.9, excessive correction easily causes a small overshoot; when it is 0.8, the optimal balance of overshoot can be achieved, effectively balancing the pre-correction accuracy and system robustness. Step S3, Overshoot Suppression and Output Calculation, introduces the amplitude and rate of change dual-correlation damping coefficient. Damping coefficient reference value The default value is 0.2, which is the maximum permissible error amplitude. Maximum permissible error rate of change Final output formula In the process, the integral term is calculated by the MCU's DMA channel in conjunction with timer interrupts. The step-by-step operation of the Simpson integral is distributed across multiple sampling cycles to reduce the workload of a single operation. The output digital control signal is converted into an analog bias voltage by a 16-bit DAC module.

[0011] Furthermore, the closed-loop optimization module is used to monitor the operating point locking accuracy, overshoot, and response time of the electro-optic modulator, and optimize the three-dimensional PID parameter mapping table in real time by combining error hysteresis rules; the control performance evaluation index includes locking accuracy. Overshoot and response time Setting error hysteresis , , Parameter updates are only performed when performance metrics exceed target values ​​and the exceedance is greater than the hysteresis, thus avoiding parameter oscillations caused by small fluctuations.

[0012] Furthermore, establish dynamic parameter update rules: if and Then increase Accelerate response and reduce Enhanced integral correction, i.e. , ;like and Then decrease Suppress overshoot and increase The smooth integral, i.e. , ;like and Then increase appropriately. , reduce ,Right now , ; Parameter constraints are , To avoid parameters exceeding a reasonable range, the update cycle is every 500ms. If the indicator meets the requirements for three consecutive times: , , If so, updates will be paused.

[0013] This invention is also achieved through the following technical solutions: A two-dimensional adaptive PID system and method for bias control of an electro-optic modulator as described above includes the following steps: Step S1: Signal Acquisition. The signal acquisition module is activated. 1% of the output feedback light from the electro-optic modulator is split by a beam splitter. After photoelectric detection, signal conditioning, and ADC conversion, three types of signals are acquired: output feedback, input optical power fluctuation, and environmental interference. The half-wave voltage is acquired simultaneously. The circuit completes the closed-loop input of the bias control signal; Step S2, Two-dimensional identification: Calculate the DC component based on the output feedback signal and slope Determine whether the operating point type is linear or extreme; calculate the PID error signal and the rate of change after filtering, and classify the error into noise, slow drift, or sudden drift error based on the dynamic threshold. Step S3, Adaptive Threshold Setting: Based on the working point type and environmental interference intensity coefficient The amplitude threshold is dynamically updated by adjusting the nonlinear adjustment coefficient. and rate of change threshold And is subject to threshold constraints, and the updated threshold is fed back to the two-dimensional recognition module; Step S4: Dual-dimensional linkage adjustment. Based on the dual-dimensional identification results and half-wave voltage, the parameters in the three-dimensional PID parameter mapping table are called according to the grading strategy. , The parameters are adjusted to perform conventional or enhanced response PI regulation; if there is an operating point switch, a three-step strategy of parameter smooth transition, error pre-correction and overshoot suppression is initiated simultaneously to output a stable bias control signal; Step S5: Closed-loop optimization. Real-time monitoring of lock-in accuracy, overshoot, and response time. Combining error hysteresis rules, determine whether to update parameters. If the update conditions are met, adjust the three-dimensional PID parameter mapping table according to dynamic rules. If the indicators meet the standards for three consecutive times, pause the update to ensure long-term system stability.

[0014] The present invention has the following beneficial effects: 1. The signal acquisition module of this invention synchronously acquires half-wave voltage. This system provides a foundation for multi-model adaptation; the dual-dimensional recognition module uses filtering to improve the accuracy of error classification, and the calibration process ensures the accuracy of slope calculation; the adaptive threshold module adapts to the nonlinear changes of environmental interference through nonlinear adjustment coefficients, resulting in stronger anti-interference capabilities; the PID adjustment module's three-dimensional parameter mapping table adapts to different models of electro-optic modulators, Simpson integral improves calculation accuracy, and dynamic switching algorithm and dual-correlation overshoot suppression ensure no overshoot during switching; the closed-loop optimization module introduces error hysteresis to avoid parameter oscillation, and dynamic ratio updates accelerate convergence; this system and method solve the problems of poor parameter adaptability, weak anti-interference, and switching overshoot in traditional PID, improves operating point locking accuracy, shortens response time, reduces the number of anti-interference misadjustments, reduces operating point switching overshoot, and reduces the bit error rate by 2-3 orders of magnitude. It is compatible with existing electro-optic modulator control frameworks, is low-cost, easy to integrate and promote, and adapts to different models of electro-optic modulators and optical communication and fiber optic sensing applications. Attached Figure Description

[0015] The present invention will now be described in further detail with reference to the accompanying drawings.

[0016] Figure 1 This is a schematic diagram of the electro-optic modulator bias control system of the present invention.

[0017] Figure 2 This is a flowchart of the method of the present invention.

[0018] Among them, 1. Laser; 2. Electro-optic modulator; 3. Coupler; 4. Beam splitter; 5. Photodetector; 6. IV preamplifier; 7. Bandpass filter; 8. Analog-to-digital converter module; 9. STM32F103RCT6; 10. Digital-to-analog converter module; 11. Signal generator; 12. Signal superposition circuit; 13. Radio frequency generator. Detailed Implementation

[0019] like Figure 1 As shown, the experimental platform is built around meeting the requirements of the software algorithm. It uses module 9 as the control core, along with basic modules 5, 6, 7, 8, and 10, to achieve a closed loop of signal acquisition, processing, and output. Module 9 needs to support timer interrupts and DMA transfer functions to ensure the efficiency of Simpson integral step-by-step calculations and real-time parameter updates. The sampling rate of module 8 is set to 102.4 kSPS, and the default signal acquisition period is 5 ms. The output range of module 10 is adapted to the bias control requirements of the electro-optic modulator. The overall hardware framework forms a closed loop. Software development is based on signal acquisition interrupt service functions, dual-dimensional recognition algorithm functions, adaptive threshold calculation functions, PID adjustment functions, closed-loop optimization functions, and operating point switching control functions. Each function uses the on-chip RAM and Flash of module 9 to achieve data interaction and parameter storage, ensuring smooth execution of the algorithm flow.

[0020] The key parameter calibration process focuses on completing the proportional constant. Configuration of the 3D PID parameter mapping table; The calibration is achieved by software controlling the output of a standard voltage signal with 0.1% accuracy from 9 to 2, while simultaneously acquiring the DC component D value of the output feedback from 2 via 8, and combining this with the measured slope of the transmission characteristic curve of 2. According to the formula Repeat the calculation 5 times and take the average to get the final result. The calibration error is 0.2%; the three-dimensional PID parameter mapping table is stored in the on-chip Flash memory of 9 according to the half-wave voltage of 2. Among them, the parameter configuration of the medium half-wave voltage range is: no adjustment for linear operating point noise error, and corresponding to slow drift error. , Sudden change drift error corresponds to , Extreme operating point noise error is not adjusted, slow drift error corresponds to... , Sudden change drift error corresponds to , Other gear parameters are configured synchronously according to the rules described in the patent, and dynamic recall based on half-wave voltage identification results is supported.

[0021] The step-by-step software execution flow of the system is as follows: The first step is signal cyclic acquisition. 9 triggers 8 to continuously acquire three types of signals at a set sampling rate via a timer. After being split by 4, the output feedback signal is conditioned by 5, 6, and 7. The input optical power fluctuation signal and environmental interference signal are also included. The output feedback signal is filtered by software to extract the DC component D and the frequency component power signal. The optical power and temperature fluctuation signals are used to calculate the environmental interference intensity coefficient. The synchronously acquired half-wave voltage data is stored in the designated RAM area of ​​9 to provide input for subsequent modules. The second step is two-dimensional recognition. The software first identifies the DC component D and the calibrated... Calculate the slope of the transmission characteristic curve 2. By matching a preset threshold The default value of 0.02 W / V is used for comparison to determine whether the operating point type is linear or extreme; then the PID error is calculated. For the target value, the error rate of change is calculated using a first-order difference combined with a three-point moving average filtering algorithm. The next step involves combining dynamic thresholds to complete error grading; the third step is to dynamically adjust the adaptive threshold according to the formula. Calculate the environmental interference intensity coefficient, and then... , Solve for the nonlinear adjustment coefficient, and call the initial threshold based on the current operating point type, according to... , The dynamic threshold is updated, and the software logic verifies whether the threshold is within the constraint range. The updated threshold is fed back to the dual-dimensional recognition module for error classification. The fourth step is dual-dimensional linkage adjustment. The software first determines the gear based on the two half-wave voltage recognition results, and then, combined with the operating point type and error characteristics, calls the corresponding value from the three-dimensional PID parameter mapping table. and The integral term is calculated using Simpson's method, and the formula is as follows: According to the adjustment formula Calculate the output bias voltage; if there is an operating point switch, the software synchronously executes a three-step strategy: parameter smooth transition according to... The transition time is calculated, and the parameters are updated according to the exponential smoothing formula. Error pre-correction is achieved by controlling 8 pre-sampling cycles (16 times) and filtering the current bias voltage 10ms before switching, and then calculating the initial error after correction. Overshoot suppression introduces a bicorrelated damping coefficient. According to the formula The calculation shows that the final output control signal, after being converted from I / O, is superimposed by the sinusoidal detection signals of I / O and I / O, and applied to the bias control terminal of I / O. The fifth step is closed-loop optimization, where the software monitors the operating point locking accuracy of I / O every 500ms. Overshoot With response time By comparing the parameter with the target value and the error hysteresis, it is determined whether to perform a parameter update: if and ,implement , ;like and ,implement , ;like and ,implement , After the parameters are updated, they are stored in the mapping table. If the indicators meet the target three times in a row, the update is paused to ensure system stability.

[0022] When writing the software, it is necessary to ensure the synchronization of the 8-sampling and 9-timer interrupts to avoid signal acquisition timing deviations; the step-by-step operation of Simpson integral needs to be distributed across multiple sampling periods to reduce the single operation load of 9; the parameter update frequency of the operating point switching needs to be consistent with the 8-sampling rate to ensure the continuity of the parameter transition curve; the hysteresis threshold of the closed-loop parameter update needs to be fine-tuned through experiments to balance the system response speed and stability.

[0023] The above description is merely a preferred embodiment of the present invention and should not be construed as limiting the scope of the present invention. All equivalent changes and modifications made in accordance with the scope of the patent application and the contents of the specification of the present invention should still fall within the scope of the patent of the present invention.

Claims

1. A two-dimensional adaptive PID system and method for bias control of an electro-optic modulator, characterized in that... The system comprises five main modules: signal acquisition, dual-dimensional recognition, adaptive thresholding, PID control, and closed-loop optimization. The signal acquisition module includes photoelectric detection, signal conditioning, and an ADC module, continuously acquiring the modulator output feedback signal, input optical power fluctuation signal, and environmental interference signal. The output feedback signal is a DC or frequency component power signal, and the environmental interference signal includes temperature fluctuations and electromagnetic interference intensity signals. The dual-dimensional recognition module determines the operating point type based on the output feedback signal and performs error classification by combining the PID error signal amplitude and the filtered rate of change. The operating point type is divided into linear and extreme operating points, and the error characteristics are divided into noise, slow drift, and abrupt drift errors. The adaptive thresholding module dynamically adjusts the error amplitude according to the operating point type and the intensity of environmental interference through a nonlinear adjustment coefficient. With the rate of change threshold The PID control module pre-stores the operating point type, error characteristics, and a three-dimensional PID parameter mapping table for half-wave voltage. Based on the identification results and the half-wave voltage, it calls up the proportional coefficient. Integral time constant The system executes a matching adjustment strategy and outputs a bias control signal. When the operating point switches, it performs a dynamic time constant parameter smooth transition, error pre-correction, and a dual correlation overshoot suppression algorithm for amplitude and rate of change. The closed-loop optimization module monitors the modulator's operating point locking accuracy, overshoot, and response time, and optimizes the three-dimensional PID parameter mapping table by combining error hysteresis rules to achieve iterative updates of PID parameters.

2. The dual-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to claim 1, characterized in that: The operating point type identification of the dual-dimensional identification module specifically involves calculating the slope of the electro-optic modulator's transmission characteristic curve based on the DC component signal D in the output feedback signal. The DC component signal D is calculated using the formula: ,in, The electro-optic conversion coefficient, For input optical power, The half-wave voltage of the electro-optic modulator. Where A is the DC bias voltage, and A is the amplitude of the sinusoidal detection signal, ranging from 1mV to 100mV. The angular frequency of the sinusoidal detection signal ranges from 1 Hz to 20 kHz. For offset phase; slope It has a linear relationship with the DC component signal D. , The proportionality constant is determined through the following calibration steps: Input a standard voltage signal at the known operating point of the electro-optic modulator, and record the corresponding D value and the measured slope. ,calculate The calibration error does not exceed 2%; the judgment rule is when... When the slope threshold is reached, it is determined to be a linear operating point. When this point is reached, it is determined to be an extreme operating point, and the slope threshold is set. The value range is 0.01~0.05W / V, with a default value of 0.02W / V.

3. The dual-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to claim 2, characterized in that: The error classification identification of the dual-dimensional identification module specifically involves calculating the PID error signal. ,in Set a value for the target. Feedback of actual values; calculation of error rate of change using first-order difference combined with 3-point moving average filtering. The approximate formula is: ,in, The signal acquisition period ranges from 1 to 10 ms, with a default value of 5 ms; the grading criterion is when... and When, it is determined to be noise error; when and When, it is determined to be a slow drift error; when and When this occurs, it is determined to be a sudden change drift error.

4. The dual-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to claim 1, characterized in that: The dynamic adjustment of the adaptive threshold module specifically involves setting the initial threshold to the initial amplitude threshold of the linear operating point. Initial rate of change threshold Extreme operating point initial amplitude threshold Initial rate of change threshold ; Define the environmental disturbance intensity coefficient ,in The input optical power fluctuation amplitude, Rated input optical power, This represents the amplitude of temperature fluctuation. Rated operating temperature; non-linear adjustment coefficient , Calculated using the following formula: , The threshold adjustment formula based on the intensity of environmental interference is as follows: , Threshold constraint is , ,in , , , .

5. The dual-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to claim 1, characterized in that, The three-dimensional PID parameter mapping rule of the PID adjustment module includes the following steps: Step S1: Using the half-wave voltage of the electro-optic modulator Based on the classification criteria, it is divided into small half-wave voltages. Medium half-wave voltage half-wave voltage Three gears, each corresponding to PID reference parameters and adjustment strategies adapted to half-wave voltage characteristics; Step S2, when At the linear operating point, if the error is noise-type interference, PID control is not executed, and the current bias is maintained; if the error exhibits slow drift characteristics, a proportional gain is used. Integral time constant Perform conventional PI control, the control formula is as follows: If the error exhibits abrupt drift characteristics, then a proportionality coefficient should be used. Integral time constant The enhanced response regulation is implemented, and the regulation formula is consistent with that of conventional PI regulation. For extreme operating points, if the error is noise-related, PID control will not be executed and the current bias will be maintained; if the error exhibits slow drift characteristics, a proportional gain will be used. Integral time constant Perform standard PI control; if the error exhibits abrupt drift characteristics, then use the proportional gain. Integral time constant Implement enhanced response adjustments; Step S3, when At the linear operating point, if the error is noise-type interference, the current bias voltage is maintained without adjustment; if the error exhibits slow drift characteristics, a proportional coefficient is used. Integral time constant If the error exhibits abrupt drift characteristics when performing conventional PI control, then the proportional gain should be used. Integral time constant Perform enhanced response adjustment; for extreme operating points, if the error is noise-type interference, maintain the current bias voltage without adjustment; if the error exhibits slow drift characteristics, use a proportional coefficient. Integral time constant If the error exhibits abrupt drift characteristics when performing conventional PI control, then the proportional gain should be used. Integral time constant When performing enhanced response adjustment, the adjustment formulas at this level are all consistent with the aforementioned conventional PI adjustment formulas. Step S4, when At the linear operating point, if the error is noise-type interference, the current bias voltage is maintained without adjustment; if the error exhibits slow drift characteristics, a proportional coefficient is used. Integral time constant If the error exhibits abrupt drift characteristics when performing conventional PI control, the proportional gain should be adjusted accordingly. Integral time constant Perform enhanced response adjustment; for extreme operating points, if the error is noise-type interference, maintain the current bias voltage without adjustment; if the error exhibits slow drift characteristics, apply a proportional coefficient. Integral time constant If the error exhibits abrupt drift characteristics when performing conventional PI control, then the proportional gain should be used. Integral time constant The enhanced response adjustment is performed, and the adjustment formula is the same as the aforementioned conventional PI adjustment formula. Step S5: All integral calculations in the above PI control process are approximated using Simpson's integral method to ensure accuracy. The approximate formula is as follows: ,in Let N be the signal acquisition period, and N be the number of integration steps, which must satisfy... .

6. The dual-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to claim 1, characterized in that, The operating point switching of the PID control module includes the following steps: Step S1: Set the parameter smooth transition, and set the switching time to... Specify the current working point parameters as follows , The target working point parameters are , Define the parameter switching range , The dynamic transition time is calculated using the following formula: The transition time ranges from 50 to 200 ms, and the formula for parameter changes with time during the transition is as follows: , ,in, ,when hour, , ; Step S2: Error pre-correction calculation, calculate the initial error correction amount before switching. ,in The target operating point bias reference value is obtained by querying the three-dimensional parameter mapping table. The current bias value; the corrected initial error is... Where 0.8 is a correction factor used to balance pre-correction accuracy and robustness; Step S3, Overshoot Suppression and Output Calculation, introduces a damping coefficient that is dual-correlated with amplitude and rate of change. The formula is ,in, This is the baseline value for the damping coefficient, ranging from 0.1 to 0.3, with a default value of 0.

2. The maximum permissible error amplitude is set to 0.5V. The maximum permissible rate of change of error is set to 0.1 V / ms; the final output formula after suppression is... .

7. The dual-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to claim 1, characterized in that: The parameter iterative update of the closed-loop optimization module specifically involves defining control performance evaluation indicators, including locking accuracy. , of which Actual operating point bias; overshoot ,in The response time is the maximum bias value during the adjustment process. From the occurrence of error to Time; setting error hysteresis , , Parameter updates are performed only when performance metrics exceed target values ​​and the exceedance is greater than the hysteresis condition; parameter update rules are established, if... and ,but , ;like and ,but , ;like and ,but , ; Parameter constraints are , ,in , , , The update cycle is once every 500ms. If the evaluation indicators meet the requirements for three consecutive times, , , If so, updates will be paused.

8. A two-dimensional adaptive PID system and method for bias control of an electro-optic modulator according to any one of claims 1-7, characterized in that: This method is applicable to optical communication and fiber optic sensing scenarios adapted to different models of electro-optic modulators. Through the coordinated operation of dual-dimensional recognition, adaptive threshold dynamic adjustment, three-dimensional PID parameter mapping, smooth switching of operating points, and closed-loop iterative optimization, it achieves high-precision and stable locking of the operating point, resulting in a technical effect of reducing the system bit error rate by 2 to 3 orders of magnitude.

Images