Digital compensation microwave frequency transfer system based on parameter optimization algorithm

Abstract

The application discloses a digital compensation microwave frequency transmission system based on a parameter optimization algorithm, which comprises a local end, an optical fiber link and a remote end, wherein the local end is connected with the remote end through the optical fiber link; a frequency signal f1 generated by the local end is modulated onto an optical carrier signal and transmitted to the remote end; a frequency signal f2 is generated by dividing the detected frequency signal f1 at the remote end; the frequency signal f2 is modulated onto another optical carrier signal and transmitted back to the local end; at the local end, a reference frequency signal f rep is measured, a phase disturbance is measured, precise compensation of the phase disturbance is realized through an electrical compensation method based on a digital phase-locked loop technology, and control parameters of the digital phase-locked loop loop are optimized in real time and on line through a parameter optimization algorithm. The application can greatly improve the performance, stability and robustness of the optical fiber microwave frequency transmission system, and can realize real-time monitoring and analysis on the system operation state.

Description

Technical Field

[0001] This invention belongs to the field of high-precision fiber optic frequency transmission, specifically relating to a digitally compensated microwave frequency transmission system based on a parameter optimization algorithm. Background Technology

[0002] With the continuous expansion of scientific frontiers and the rapid development of engineering technology, the performance requirements of high-precision frequency signals are constantly increasing, and their application fields are becoming increasingly widespread. Therefore, utilizing high-precision frequency transmission technology to realize the long-distance networked transmission of standard frequency signals is an important way to meet the high-precision frequency signal requirements of fields such as time and frequency metrology, basic scientific research, and quantum information network construction.

[0003] Compared with traditional frequency transmission methods, fiber-optic microwave frequency transmission technology has outstanding advantages such as high precision, low loss, low added noise, and abundant channel resources. Currently, fiber-optic microwave transmission systems utilize analog phase-locked loop (PLL) technology to precisely compensate for phase disturbances (such as temperature, humidity, stress, and vibration) experienced by standard frequency signals during transmission through fiber optic links, thereby achieving high-precision microwave frequency transmission.

[0004] However, fiber optic microwave frequency transfer technology based on analog phase-locked loops can only set fixed loop parameters, making it difficult to adapt to various disturbances and changes in environmental conditions in the transmission link. At the same time, the analog system also has difficulty in recording, analyzing, and uploading various system operating parameters. Summary of the Invention

[0005] To address the aforementioned problems in the existing technology, this invention provides a digitally compensated microwave frequency transfer system based on a parameter optimization algorithm. The technical problem to be solved by this invention is achieved through the following technical solution:

[0006] This invention provides a digital compensation microwave frequency transfer system based on a parameter optimization algorithm. The system includes a local end, an optical fiber link, and a remote end, with the local end and the remote end connected via an optical fiber link.

[0007] The local unit includes a first power divider, a second power divider, a first mixer, a second mixer, a first N-dividend, a voltage-controlled oscillator, a digital phase detector, an analog-to-digital converter, a processor, a digital-to-analog converter, a first laser, a first modulator, and a first optical path detector; among which,

[0008] The first power divider is used to divide the reference frequency signal f rep The circuit is divided into two paths; the voltage-controlled oscillator is used to generate a frequency signal f1 based on an initially set frequency signal; the second power divider is used to divide the frequency signal f1 into two paths; the first mixer is used to divide the frequency signal f1 into two paths based on a reference frequency signal f1. repA first frequency signal is generated by mixing a frequency signal f1 with another frequency signal f1; a first N-divider is used to generate a first frequency-divided signal based on the first mixing signal; a first laser is used to generate a first optical carrier signal; a first modulator is used to generate a first modulation signal based on the first optical carrier signal and another frequency signal f1, so as to generate a frequency signal f2 at the remote end that returns to the local end; wherein, the frequency signal f2 is a frequency-divided signal of the frequency signal f1 detected at the remote end and carries phase disturbances generated during optical fiber link transmission; a second mixer is used to generate a first mixing signal based on another reference frequency signal f1. rep The first frequency division signal and the second frequency mixing signal are used to generate a second mixing signal; the digital phase detector is used to generate an error voltage signal based on the first frequency division signal and the second mixing signal; the analog-to-digital converter is used to digitize the error voltage signal; the processor is used to continuously search for optimal control parameters online using a parameter optimization algorithm and calculate the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal; the digital-to-analog converter is used to analogize the phase-compensated voltage signal; the voltage-controlled oscillator is also used to generate a new frequency signal f1 based on the analog voltage signal.

[0009] In one embodiment of the present invention, the remote end includes a first photodetector, a second laser, a second modulator, a third power divider, a second N-divider, and a frequency synthesizer; wherein,

[0010] The first photodetector is used to detect the frequency signal f1 from the first modulation signal; the third power divider is used to split the frequency signal f1 into two paths; the second N-divider is used to generate a frequency signal f2 based on one path of the frequency signal f1; the second laser is used to generate a second optical carrier signal; the second modulator is used to generate a second modulation signal based on the second optical carrier signal and the frequency signal f2; and the frequency synthesizer is used to generate the reference frequency signal f1 based on the other path of the frequency signal f1. rep .

[0011] In one embodiment of the present invention, the local terminal further includes a first optical circulator and a second photodetector; wherein...

[0012] The first optical circulator is used to control the flow of the second modulation signal to the second photodetector; the second photodetector is used to detect the frequency signal f2 from the second modulation signal.

[0013] In one embodiment of the present invention, the remote end further includes a second optical circulator; the second optical circulator is used to control the flow of the first modulation signal to the first photodetector.

[0014] In one embodiment of the present invention, the initially set frequency signal in the voltage-controlled oscillator is a reference frequency signal f. rep (N+1) / 2 times, where N represents the division coefficient of the first N-divider.

[0015] In one embodiment of the present invention, the processor uses a parameter optimization algorithm to continuously search for and optimize control parameters online, and calculates the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal, as expressed by the formula:

[0016]

[0017] Where k represents the number of samples, V f (k) represents the phase-compensated voltage signal output by the processor at the k-th sampling time, V e (k) represents the digitized error voltage signal at the k-th sampling time, V e (k-1) represents the digitized error voltage signal at the (k-1)th sampling time, K p K represents the proportional control parameter. i K represents the integral control parameter. d K represents the differential control parameter. p K i K d This is to obtain the control parameters that need to be continuously searched and optimized online using a parameter optimization algorithm.

[0018] In one embodiment of the present invention, the parameter optimization algorithm in the processor includes the Fibonacci method, the golden section method, the direction acceleration method, and the simplex acceleration method.

[0019] In one embodiment of the present invention, if the parameter optimization algorithm in the processor is a simplex acceleration method, the process of continuously searching for and optimizing control parameters online using the parameter optimization algorithm and calculating the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal includes:

[0020] Based on the digitized error voltage signal, a target optimization function controlled by the control parameters is constructed.

[0021] Initialize three sets of control parameters;

[0022] Based on the three sets of control parameters, the corresponding objective function values ​​are calculated and compared according to the objective optimization function. The point where the control parameter with the smallest objective function value is located is taken as the starting point X. L The point containing the control parameter with the larger objective function value is taken as the starting point X. G The point containing the control parameter with the largest objective function value is taken as the starting point X. H ;

[0023] Based on the starting point X G Starting point X L Starting point X H The simplex {X} formed G X L X M The system continuously searches online for the control parameters that minimize the objective function value of the target optimization function, and uses these as the optimal control parameters to calculate the phase-compensated voltage signal.

[0024] In one embodiment of the present invention, the constructed objective optimization function is expressed as follows:

[0025]

[0026] Where X represents the control parameter, f(X) represents the objective function value corresponding to the control parameter X, and V e (j) represents the digitized error voltage signal at the j-th sampling time, and k represents the number of samplings.

[0027] In one embodiment of the invention, based on the starting point X G Starting point X L Starting point X H The simplex {X} formed G X L X M The process involves continuously searching online for the control parameters that minimize the objective function value of the target optimization function, and using these as the optimal control parameters.

[0028] Take the starting point X L and the starting point X G Midpoint X of the line F Connect the midpoint X F and the starting point X H and from the midpoint X F Extend to reflection point X R And according to the reflection point X R Calculate the corresponding objective function value for the corresponding control parameters; where the extension length is the midpoint X. F and the starting point X H The length of the connecting line;

[0029] Determine if f(X) R ) <f(X L Then from the reflection point X R Continue extending to expansion point X E Point X E -X F =α(X) F -X H), where α represents the expansion factor; if f(X) E ) <f(X R Then, the expansion point X E Instead of the starting point X H To form a new simplex {X G X L X E}, return to the steps of calculating and comparing the corresponding objective function values ​​based on the three sets of control parameters and the objective optimization function; if f(X) E )>f(X R Then, taking the reflection point X as an example... R Instead of the starting point X H To form a new simplex {X G X L X R}, set the starting point X G Starting point X L Reflection point X R As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned.

[0030] Determine if f(X) L ) <f(X R ) <f(X G Then, taking the reflection point X as an example... R Instead of the starting point X H To form a new simplex {X G X L X R}, set the starting point X G Starting point X L Reflection point X R As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned.

[0031] Determine if f(X) G ) <f(X R ) <f(X H Then from the midpoint X F Extend to compression point X S , making X S -X F =β(X) F -X H ), where β represents the compression factor, and the compression point X is... S Instead of the starting point X H To form a new simplex {X G X L X S}, set the starting point X G Starting point X L Compression point X S As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned.

[0032] Determine if f(X) H ) <f(X R ), then from the starting point X H Extend to compression point X M , making X F -X M =β(X) F -X H If f(X) M ) <f(X H Then, the compression point X M Instead of the starting point X H To form a new simplex {X G X L X M}, set the starting point X G Starting point X L Compression point X M As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned; if f(X) H ) <f(X M If so, then take the starting point X. H The starting point X L The midpoint X N The midpoint X F The midpoint X N Instead of the starting point X H The starting point X G To form a new simplex {X F X L X N}, place the midpoint X F Starting point X L Midpoint X N As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned.

[0033] Determine if the search count has reached the maximum search count and if the optimization result meets the optimization accuracy condition. If the search count has not reached the maximum search count and the optimization result does not meet the optimization accuracy condition, then return a value based on the starting point X. G Starting point X L Starting point X HThe simplex {X} formed G X L X M The process involves continuously searching online for the control parameter that minimizes the objective function value of the target optimization function. If the number of searches has not reached the maximum and the optimization result meets the optimization accuracy condition, the optimization search is successful. The control parameter that minimizes the objective function value of the target optimization function is selected from the current simplex as the optimal control parameter. If the number of searches has reached the maximum and the optimization result does not meet the optimization accuracy condition, the optimization search fails, and the optimal control parameter output from the previous search is retained.

[0034] The beneficial effects of this invention are:

[0035] The proposed digital compensation microwave frequency transfer system based on a parameter optimization algorithm modulates a frequency signal f1 generated at the local end onto an optical carrier signal of one channel and transmits it to a remote end via an optical fiber link. At the remote end, the detected frequency signal f1 is divided to generate a frequency signal f2, which carries phase disturbances generated during transmission via the optical fiber link. This frequency signal f2 is then modulated onto an optical carrier signal of another channel and transmitted back to the local end. At the local end, the reference frequency signal f1 is then processed... rep Digital phase detection is performed on the returned frequency signal f2 to measure the phase disturbance introduced by changes in the external environment (such as temperature, humidity, stress, vibration, etc.) of the fiber optic link. Based on digital phase-locked loop (PLL) technology, electrical compensation methods are used to accurately compensate for the phase disturbance. Simultaneously, a parameter optimization algorithm is used to optimize the control parameters of the PLL loop in real time to adapt to changes in environmental conditions and system disturbances. It is evident that by leveraging the inherent advantages of PLL technology and combining it with parameter optimization algorithms for real-time optimization of the PLL loop control parameters, compared to traditional analog PLL technology, the performance, stability, and robustness of fiber optic microwave frequency transmission systems can be significantly improved. Furthermore, PLL technology provides favorable conditions for the quantification, analysis, storage, and uploading of parameters such as phase detection signals and control signals, facilitating real-time monitoring and analysis of system operation status, and thus enabling better application in the field of fiber optic frequency transmission.

[0036] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0037] Figure 1 This is a schematic diagram of the structure of a digital compensation microwave frequency transfer system based on a parameter optimization algorithm provided in an embodiment of the present invention;

[0038] Figure 2 This is a schematic diagram of a search performed on a PI two-dimensional plane according to an embodiment of the present invention;

[0039] Figure 3 This is a schematic diagram of the process of continuously searching for and optimizing control parameters online using a parameter optimization algorithm, as provided in an embodiment of the present invention. Detailed Implementation

[0040] The present invention will be further described in detail below with reference to specific embodiments, but the implementation of the present invention is not limited thereto.

[0041] Please see Figure 1 This invention provides a digital compensation microwave frequency transfer system based on a parameter optimization algorithm. The system includes a local end, an optical fiber link, and a remote end, with the local end and the remote end connected by an optical fiber link.

[0042] The local unit includes a first power divider, a second power divider, a first mixer, a second mixer, a first N-dividend, a voltage-controlled oscillator, a digital phase detector, an analog-to-digital converter, a processor, a digital-to-analog converter, a first laser, a first modulator, and a first optical path detector; among which,

[0043] The first power divider is used to divide the reference frequency signal f rep It is divided into two paths: a voltage-controlled oscillator (VCO) to generate a frequency signal f1 based on an initially set frequency signal; a second power divider to split the frequency signal f1 into two paths; and a first mixer to generate a frequency signal f1 based on a reference frequency signal f1. rep A first mixer is used to generate a first mixed signal based on a frequency signal f1; a first N-divider is used to generate a first divided signal based on the first mixed signal; a first laser is used to generate a first optical carrier signal; a first modulator is used to generate a first modulated signal based on the first optical carrier signal and another frequency signal f1, so as to generate a frequency signal f2 at the remote end that returns to the local end; wherein, the frequency signal f2 is a divided signal of the frequency signal f1 detected at the remote end and carries phase disturbances generated during optical fiber link transmission; a second mixer is used to generate a first mixed signal based on another reference frequency signal f1. rep A second mixing signal is generated from the frequency signal f2; a digital phase detector is used to generate an error voltage signal based on the first frequency division signal and the second mixing signal; an analog-to-digital converter is used to digitize the error voltage signal; a processor is used to continuously search for optimal control parameters online using a parameter optimization algorithm and to calculate the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal; a digital-to-analog converter is used to analogize the phase-compensated voltage signal; and a voltage-controlled oscillator is also used to generate a new frequency signal f1 based on the analog voltage signal.

[0044] In this embodiment of the invention, the remote end includes a first photodetector, a second laser, a second modulator, a third power divider, a second N-divider, and a frequency synthesizer; wherein,

[0045] A first photodetector is used to detect the frequency signal f1 from the first modulation signal; a third power divider is used to split the frequency signal f1 into two paths; a second N-divider is used to generate a frequency signal f2 based on one path of the frequency signal f1; a second laser is used to generate a second optical carrier signal; a second modulator is used to generate a second modulation signal based on the second optical carrier signal and the frequency signal f2; and a frequency synthesizer is used to generate a reference frequency signal f1 based on the other path of the frequency signal f1. rep Among them, the first power divider, the third power divider, and the second power divider correspond to two signal channels with the same frequency, and the power can be the same or different, depending on the actual situation.

[0046] In this embodiment of the invention, the local terminal further includes a first optical circulator and a second photodetector; wherein,

[0047] The first optical looper is used to control the flow of the second modulation signal to the second photodetector; the second photodetector is used to detect the frequency signal f2 from the second modulation signal.

[0048] In this embodiment of the invention, the remote end further includes a second optical circulator; the second optical circulator is used to control the flow of the first modulation signal to the first photodetector.

[0049] In this embodiment of the invention, the initially set frequency signal in the voltage-controlled oscillator is the reference frequency signal f. rep The division factor is (N+1) / 2 times, where N represents the division coefficient of the first N-divider. The division coefficient of the first N-divider is the same as that of the second N-divider. The specific division coefficient is selected according to the actual situation. For example, a divider with a 2-dividend ratio can be selected. The signal power after division is the same, but the frequency can be the same or different.

[0050] As can be seen, at the local end, the voltage-controlled oscillator generates a frequency signal f1, the frequency of which is set to the reference signal f. rep The frequency signal f1, multiplied by (N+1) / 2, is split into two paths by the second power divider: one path modulates the frequency signal f1 onto the optical carrier signal generated by the first laser via the first modulator, and transmits it to the remote end via an optical fiber link. At the remote end, it is detected by the first photodetector and then processed by the second N-divider to generate a frequency signal f2 = f1 / N, which is then transmitted back to the local end and detected by the second photodetector. At this point, the frequency signal f2 transmitted back to the local end carries a phase disturbance Δθ caused by the optical fiber link. This transmitted frequency signal f2 differs from the reference frequency signal f1. rep The second mixer generates the second mixing signal (f). rep -f2-Δθ); another frequency signal f1 and the reference frequency signal f rep The first mixer generates the first mixing signal (f2-f). repThen, the first frequency divider (N) generates the first frequency divided signal (f2-f). rep ) / N. The first frequency division signal (f2-f rep ) / N and the second mixer signal (f rep The phase disturbance Δθ is fed into a digital phase detector, where it is converted into an error voltage signal V. e (t), the error voltage signal V e (t) After being digitized by an analog-to-digital converter, the voltage signal V is acquired by a processor and calculated using a parameter optimization algorithm to generate a phase-compensated voltage signal. f (k) will convert the voltage signal V f (k) After being analogized by a digital-to-analog converter, it is input to the voltage control terminal of the voltage-controlled oscillator to achieve accurate compensation for the phase disturbance Δθ of the optical fiber link.

[0051] At the remote end, the frequency signal f1 transmitted from the local end is detected by the first photodetector and split into two paths by the third power divider: one path, frequency signal f1, is directly divided by the second N-divider to generate frequency signal f2. Frequency signal f2 is modulated by the second modulator onto the optical carrier signal generated by the second laser and transmitted back to the local end via the same optical fiber link. This signal carries the phase disturbance Δθ caused in the optical fiber link, providing a reference for accurate compensation at the local end; the other path, frequency signal f1, is used by the frequency synthesizer to generate a reference frequency signal f. rep Because the frequency signal f1 at the local end has undergone phase compensation, the reference frequency signal f1 recovered from the frequency signal f1 at the remote end is... rep The phase of the local reference frequency signal is almost perfectly synchronized, thus achieving the reference frequency signal f. rep It is transmitted from the local end to the remote end via a fiber optic link.

[0052] Furthermore, this embodiment of the invention employs digital phase-locked loop (PLL) technology. This is because PLLs naturally facilitate the recording, analysis, and uploading of various system operating parameters, enhancing data analysis and real-time monitoring of system operation. Moreover, PLL technology is easily integrated with various algorithms to set optimal loop parameters for various link disturbances and environmental condition changes, thereby improving system performance, stability, and robustness. As is well known, the loop parameters of a PLL, such as loop gain, damping coefficient, and loop bandwidth, are all determined by the transfer functions of each device in the loop and K... p K i K d These three control parameters are determined. Theoretically, when other parameters of the phase-locked loop are determined, a suitable control parameter K can be selected. p K i K d This can achieve good phase compensation. However, in practical engineering applications, the control parameter K... pK i K d The adaptability to changes in system characteristics is limited, with a fixed control parameter K. p K i K d It is difficult to adapt to changes in environmental conditions and system disturbances, therefore the control parameter K is... p K i K d Online adaptive optimization search is essential.

[0053] To address the aforementioned needs, this invention proposes using a parameter optimization algorithm within the processor to continuously search for and optimize control parameters online. Based on these control parameters and the digitized error voltage signal, the phase-compensated voltage signal is calculated. The formula is as follows:

[0054]

[0055] Where k represents the number of samples, V f (k) represents the phase-compensated voltage signal output by the processor at the k-th sampling time, V e (k) represents the digitized error voltage signal at the k-th sampling time, V e (k-1) represents the digitized error voltage signal at the (k-1)th sampling time, K p The proportional control parameter is denoted as P parameter, K. i The integral control parameter is denoted as I parameter, K. d The differential control parameters are denoted as D parameters, K. p K i K d The control parameters need to be continuously searched and optimized online using a parameter optimization algorithm. As can be seen from formula (1), the control parameters K obtained through online search optimization... p K i K d It can achieve phase compensation better.

[0056] The parameter optimization algorithms in the processor of this invention include the Fibonacci method, the golden section method, the directional acceleration method, and the simplex acceleration method. Taking the simplex acceleration method as an example, the control parameter K is described in detail. p K i K d The search and optimization process. Specifically:

[0057] In this embodiment of the invention, when the parameter optimization algorithm in the processor is the simplex acceleration method, the corresponding process of continuously searching for and optimizing control parameters online using the parameter optimization algorithm and calculating the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal includes:

[0058] Based on the digitized error voltage signal, a target optimization function controlled by the control parameters is constructed; three sets of control parameters are initialized; based on the three sets of control parameters, the corresponding objective function values ​​are calculated and compared, and the point containing the control parameter with the smallest objective function value is taken as the starting point X. L The point containing the control parameter with the larger objective function value is taken as the starting point X. G The point containing the control parameter with the largest objective function value is taken as the starting point X. H Based on the starting point X G Starting point X L Starting point X H The simplex {X} formed G X L X M The system continuously searches online for the control parameters that minimize the objective function value of the target optimization function. These parameters are then used as the optimal control parameters, and the phase-compensated voltage signal is calculated based on these optimal control parameters.

[0059] This invention uses the output error voltage signal to determine the performance of the control parameters and employs the control parameter with the minimum ISTE (Integral Square Time Square Error) to adapt to changes in environmental conditions and system disturbances. Therefore, the constructed objective optimization function is expressed as follows:

[0060]

[0061] Where X represents the control parameter, f(X) represents the objective function value corresponding to the control parameter X, and V e (j) represents the digitized error voltage signal at the j-th sampling time, and k represents the number of samplings. The smaller the objective function value f(X) of the objective optimization function shown in formula (2), the better the control effect of the corresponding control parameter X.

[0062] To better understand, assume that the P parameters and I parameters together form a two-dimensional simplex as follows: Figure 2 As shown, each set of PI parameters forms a point on a two-dimensional plane. Based on experience, three sets of control parameters X are initialized. H X G X L These three vertices form a triangle on a two-dimensional plane defined by the P and I parameters. See also... Figure 3In the process of continuously searching and optimizing control parameters online in this embodiment of the invention, α, β, and ε are first given, K = 0, α represents the expansion factor, and α takes any value greater than 1, β represents the compression factor, and β takes a value from 0 to 1, K represents the number of searches, and ε represents the optimization accuracy, and ε takes a value from 0 to 1. The maximum number of searches is K. max Specifically, α, β, ε, K max It is set according to the actual situation;

[0063] Calculate the values ​​at the starting point X according to formula (2). G Starting point X L Starting point X H The objective function value f(X) at point H f(X) G f(X) L ), and compare their sizes, assuming f(X) H )>f(X G )>f(X L ), indicating the starting point X H The worst performance, starting point X G The performance is second best, starting point X L It has the best performance.

[0064] This invention embodiment is based on starting point X G Starting point X L Starting point X H The simplex {X} formed G X L X M The system continuously searches online for the control parameters that minimize the objective function value of the objective function, and uses these as the optimal control parameters.

[0065] Take the starting point X L and starting point X G Midpoint X of the line F Connect the midpoint X F and starting point X H and from the midpoint X F Extend to reflection point X R And based on the reflection point X R Calculate the corresponding objective function value f(X) for the corresponding control parameters. R ); where the extension length is the midpoint X F and starting point X H The length of the connection; that is, from the initial point X with the worst performance. H Start the search.

[0066] For calculating f(X) R After that, the following situations may occur:

[0067] Determine if f(X) R ) <f(X L This indicates that the search direction is correct, and we can continue searching along the starting point X. H Reflection point X R The directional expansion search starts from the reflection point X. R Continue extending to expansion point X E , making X E -X F =α(X) F -X H ), and calculate the expansion point X. E The corresponding objective function value f(X) E If f(X) E ) <f(X R This indicates that expansion is favorable, so the expansion point X is taken as the starting point. E Replace the starting point X H To form a new simplex {X G X L X E} Proceed to the next search, that is, return to the step of calculating and comparing the corresponding objective function values ​​based on the three sets of control parameters and the objective optimization function; if f(X) E )>f(X R This indicates that expansion is unfavorable, so the reflection point X is taken as the reference point. R Replace the starting point X H To form a new simplex {X G X L X R}, set the starting point X G Starting point X L Reflection point X R The next search is performed using the three new sets of control parameters, which means returning to the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters.

[0068] Determine if f(X) L ) <f(X R ) <f(X G This indicates that the search direction is correct and no expansion is needed. Therefore, the reflection point X is used as the reference point. R Replace the starting point X H To form a new simplex {X G X L X R}, set the starting point X G Starting point X L Reflection point X R The next search is performed using the three new sets of control parameters, which means returning to the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters.

[0069] Determine if f(X) G ) <f(X R ) <f(X H This indicates that at the starting point X... H The search step size along the reflection point direction is too long; it should be reduced. Therefore, the search step size should start from the midpoint X. F Extend to compression point X S , making X S -X F =β(X) F -X H ), and with compression point X S Replace the starting point X H To form a new simplex {X G X L X S}, set the starting point X G Starting point X L Compression point X S The next search is performed using the three new sets of control parameters, which means returning to the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters.

[0070] Determine if f(X) H ) <f(X R This indicates that at the starting point X... H The search compensation along the reflection point direction should be further compressed, starting from the initial point X. H Extend to compression point X M , making X F -X M =β(X) F -X H ), and calculate the compression point X. M The corresponding objective function value f(X) M If f(X) M ) <f(X H Then, the compression point X M Replace the starting point X H To form a new simplex {X G X L X M}, set the starting point X G Starting point X L Compression point X M The next search will use the three new sets of control parameters as the basis for the new set of control parameters, which means returning to the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters; if f(X) H ) <f(X M This indicates that at the starting point X... H All combinations of control parameters in the direction of the reflection point are inferior to those at the starting point X. HTherefore, the search cannot proceed in that direction; the starting point should be X. L The simplex {X} centered at the center H X G X L Perform contraction and take the starting point X. H Starting point X L The midpoint X N Midpoint X F Midpoint X N Replace the starting point X H Starting point X G To form a new simplex {X F X L X N}, place the midpoint X F Starting point X L Midpoint X N The next search is performed using the three new sets of control parameters, which means returning to the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters.

[0071] Determine if the number of searches has reached the maximum number of searches K. max Furthermore, it is determined whether the optimization result satisfies the optimization accuracy condition, which is defined as f(X) H )-f(X L )<εf(X L If the number of searches does not reach the maximum number of searches and the optimization result does not meet the optimization accuracy condition, then return a result based on the starting point X. G Starting point X L Starting point X H The simplex {X} formed G X L X M The process involves continuously searching online for the control parameter that minimizes the objective function value of the target optimization function. If the number of searches has not reached the maximum and the optimization result meets the optimization accuracy condition, the optimization search is successful, and the control parameter that minimizes the objective function value of the target optimization function is selected from the current simplex as the optimal control parameter. If the number of searches has reached the maximum and the optimization result does not meet the optimization accuracy condition, the optimization search fails, and the optimal control parameter output from the previous search is retained.

[0072] In summary, the digital compensation microwave frequency transfer system based on parameter optimization algorithm proposed in this invention modulates the frequency signal f1 generated at the local end onto an optical carrier signal of one channel and transmits it to the remote end via an optical fiber link. At the remote end, the detected frequency signal f1 is divided to generate a frequency signal f2. The frequency signal f2 carries the phase disturbance generated during transmission via the optical fiber link. The frequency signal f2 is modulated onto an optical carrier signal of another channel and transmitted back to the local end. At the local end, the reference frequency signal f1 is then processed... rep Digital phase detection is performed on the returned frequency signal f2 to measure the phase disturbance introduced by changes in the external environment (such as temperature, humidity, stress, vibration, etc.) of the fiber optic link. Based on digital phase-locked loop (PLL) technology, electrical compensation methods are used to accurately compensate for the phase disturbance. Simultaneously, a parameter optimization algorithm is used to optimize the control parameters of the PLL loop in real time to adapt to changes in environmental conditions and system disturbances. It is evident that by leveraging the inherent advantages of PLL technology and combining it with parameter optimization algorithms for real-time optimization of the PLL loop control parameters, compared to traditional analog PLL technology, the performance, stability, and robustness of fiber optic microwave frequency transmission systems can be significantly improved. Furthermore, PLL technology provides favorable conditions for the quantification, analysis, storage, and uploading of parameters such as phase detection signals and control signals, facilitating real-time monitoring and analysis of system operation status, and thus enabling better application in the field of fiber optic frequency transmission.

[0073] In the description of this invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0074] Although the invention has been described herein in conjunction with various embodiments, those skilled in the art, by reviewing the specification and accompanying drawings, will understand and implement other variations of the disclosed embodiments in carrying out the claimed invention. In the specification, the word "comprising" does not exclude other components or steps, and "a" or "an" does not exclude a plurality. While certain measures are described in different embodiments, this does not mean that these measures cannot be combined to produce good results.

[0075] The above description, in conjunction with specific preferred embodiments, provides a further detailed explanation of the present invention. It should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various simple deductions or substitutions can be made without departing from the concept of the present invention, and all such modifications and substitutions should be considered within the scope of protection of the present invention.

Claims

1. A digitally compensated microwave frequency transfer system based on a parameter optimization algorithm, characterized in that, The system includes a local end, an optical fiber link, and a remote end, with the local end and the remote end connected via an optical fiber link; The local unit includes a first power divider, a second power divider, a first mixer, a second mixer, a first N-dividend, a voltage-controlled oscillator, a digital phase detector, an analog-to-digital converter, a processor, a digital-to-analog converter, a first laser, a first modulator, and a first optical path detector; among which, The first power divider is used to divide the reference frequency signal f rep The circuit is divided into two paths; the voltage-controlled oscillator is used to generate a frequency signal f1 based on an initially set frequency signal; the second power divider is used to divide the frequency signal f1 into two paths; the first mixer is used to generate a frequency signal f1 based on a reference frequency signal f1. rep A first frequency signal is generated by mixing a frequency signal f1 with another frequency signal f1; a first N-divider is used to generate a first frequency-divided signal based on the first mixing signal; a first laser is used to generate a first optical carrier signal; a first modulator is used to generate a first modulation signal based on the first optical carrier signal and another frequency signal f1, so as to generate a frequency signal f2 at the remote end that returns to the local end; wherein, the frequency signal f2 is a frequency-divided signal of the frequency signal f1 detected at the remote end and carries phase disturbances generated during optical fiber link transmission; a second mixer is used to generate a first mixing signal based on another reference frequency signal f1. rep The first frequency division signal and the second frequency mixing signal are used to generate a second mixing signal; the digital phase detector is used to generate an error voltage signal based on the first frequency division signal and the second mixing signal; the analog-to-digital converter is used to digitize the error voltage signal; the processor is used to continuously search for optimal control parameters online using a parameter optimization algorithm and calculate the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal; the digital-to-analog converter is used to analogize the phase-compensated voltage signal; the voltage-controlled oscillator is also used to generate a new frequency signal f1 based on the analog voltage signal.

2. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 1, characterized in that, The remote end includes a first photodetector, a second laser, a second modulator, a third power divider, a second N-divider, and a frequency synthesizer; among which, The first photodetector is used to detect the frequency signal f1 from the first modulation signal; the third power divider is used to split the frequency signal f1 into two paths; the second N-divider is used to generate a frequency signal f2 based on one path of the frequency signal f1; the second laser is used to generate a second optical carrier signal; the second modulator is used to generate a second modulation signal based on the second optical carrier signal and the frequency signal f2; and the frequency synthesizer is used to generate the reference frequency signal f1 based on the other path of the frequency signal f1. rep .

3. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 2, characterized in that, The local end also includes a first optical circulator and a second photodetector; among which, The first optical circulator is used to control the flow of the second modulation signal to the second photodetector; the second photodetector is used to detect the frequency signal f2 from the second modulation signal.

4. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 2, characterized in that, The remote end also includes a second optical circulator; the second optical circulator is used to control the flow of the first modulation signal to the first photodetector.

5. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 1, characterized in that, The initial frequency signal set in the voltage-controlled oscillator is the reference frequency signal f. rep (N+1) / 2 times; where N represents the division coefficient of the first N-divider.

6. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 1, characterized in that, The processor uses a parameter optimization algorithm to continuously search for and optimize control parameters online, and calculates the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal. The formula is as follows: Where k represents the number of samples, V f (k) represents the phase-compensated voltage signal output by the processor at the k-th sampling time, V e (k) represents the digitized error voltage signal at the k-th sampling time, V e (k-1) represents the digitized error voltage signal at the (k-1)th sampling time, K p K represents the proportional control parameter. i K represents the integral control parameter. d K represents the differential control parameter. p K i K d This is to obtain the control parameters that need to be continuously searched and optimized online using a parameter optimization algorithm.

7. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 1, characterized in that, The parameter optimization algorithms in the processor include the Fibonacci method, the golden section method, the direction acceleration method, and the simplex acceleration method.

8. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 7, characterized in that, If the parameter optimization algorithm in the processor is the simplex acceleration method, the corresponding process of continuously searching for and optimizing control parameters online using the parameter optimization algorithm, and calculating the phase-compensated voltage signal based on the control parameters and the digitized error voltage signal, includes: Based on the digitized error voltage signal, a target optimization function controlled by the control parameters is constructed. Initialize three sets of control parameters; Based on the three sets of control parameters, the corresponding objective function values ​​are calculated and compared according to the objective optimization function. The point where the control parameter with the smallest objective function value is located is taken as the starting point X. L The point containing the control parameter with the larger objective function value is taken as the starting point X. G The point containing the control parameter with the largest objective function value is taken as the starting point X. H ; Based on the starting point X G Starting point X L Starting point X H The simplex {X} formed G X L X M The system continuously searches online for the control parameters that minimize the objective function value of the target optimization function, and uses these as the optimal control parameters to calculate the phase-compensated voltage signal.

9. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 8, characterized in that, The constructed objective optimization function is expressed as follows: Where X represents the control parameter, f(X) represents the objective function value corresponding to the control parameter X, and V e (j) represents the digitized error voltage signal at the j-th sampling time, and k represents the number of samplings.

10. The digital compensation microwave frequency transfer system based on parameter optimization algorithm according to claim 9, characterized in that, Based on the starting point X G Starting point X L Starting point X H The simplex {X} formed G X L X M The process involves continuously searching online for the control parameters that minimize the objective function value of the target optimization function, and using these as the optimal control parameters. Take the starting point X L and the starting point X G Midpoint X of the line F Connect the midpoint X F and the starting point X H and from the midpoint X F Extend to reflection point X R And according to the reflection point X R Calculate the corresponding objective function value for the corresponding control parameters; where the extension length is the midpoint X. F and the starting point X H The length of the connecting line; Determine if f(X) R ) <f(X L Then from the reflection point X R Continue extending to expansion point X E Point X E -X F =α(X) F -X H ), where α represents the expansion factor; if f(X) E ) <f(X R Then, with the expansion point X E Instead of the starting point X H To form a new simplex {X G X L X E }, return to the steps of calculating and comparing the corresponding objective function values ​​based on the three sets of control parameters and the objective optimization function; if f(X) E )>f(X R Then, taking the reflection point X as an example... R Instead of the starting point X H To form a new simplex {X G X L X R }, set the starting point X G Starting point X L Reflection point X R As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned. Determine if f(X) L ) <f(X R ) <f(X G Then, taking the reflection point X as an example... R Instead of the starting point X H To form a new simplex {X G X L X R }, set the starting point X G Starting point X L Reflection point X R As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned. Determine if f(X) G ) <f(X R ) <f(X H Then from the midpoint X F Extend to compression point X S , making X S -X F =β(X) F -X H ), where β represents the compression factor, and the compression point X is... S Instead of the starting point X H To form a new simplex {X G X L X S }, set the starting point X G Starting point X L Compression point X S As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned. Determine if f(X) H ) <f(X R ), then from the starting point X H Extend to compression point X M , making X F -X M =β(X) F -X H If f(X) M ) <f(X H Then, the compression point X M Instead of the starting point X H To form a new simplex {X G X L X M }, set the starting point X G Starting point X L Compression point X M As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned; if f(X) H ) <f(X M If so, then take the starting point X. H The starting point X L midpoint X N The midpoint X F The midpoint X N Instead of the starting point X H The starting point X G To form a new simplex {X F X L X N }, place the midpoint X F Starting point X L Midpoint X N As three new sets of control parameters, the steps of calculating and comparing the corresponding objective function values ​​based on the objective optimization function according to the three sets of control parameters are returned. Determine if the search count has reached the maximum search count and if the optimization result meets the optimization accuracy condition. If the search count has not reached the maximum search count and the optimization result does not meet the optimization accuracy condition, then return a value based on the starting point X. G Starting point X L Starting point X H The simplex {X} formed G X L X M The process involves continuously searching online for the control parameter that minimizes the objective function value of the target optimization function. If the number of searches has not reached the maximum and the optimization result meets the optimization accuracy condition, the optimization search is successful. The control parameter that minimizes the objective function value of the target optimization function is selected from the current simplex as the optimal control parameter. If the number of searches has reached the maximum and the optimization result does not meet the optimization accuracy condition, the optimization search fails, and the optimal control parameter output from the previous search is retained.

Images