Parameter determination method, device and equipment of distributed Raman optical fiber amplifier and medium

Abstract

The invention discloses a parameter determination method, device and equipment for a distributed Raman optical fiber amplifier and a medium, and the method comprises the steps: determining an optimization objective function of link tail end power configuration of each wavelength channel based on a target application scene of an optical fiber sensing integrated system; solving the optimization objective function by using a simulated annealing algorithm to obtain the optimal link end power configuration of each wavelength channel; and obtaining the target pumping power of each pumping source in the distributed Raman optical fiber amplifier by using a pre-trained neural network model according to the incident fiber optical power of each wavelength channel and the optimal link end power configuration. According to the invention, different requirements of communication and sensing can be met at the same time, and the communication and sensing fusion gain is maximized.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a method, device, electronic equipment and computer-readable storage medium for determining parameters of a distributed Raman fiber amplifier. Background Art

[0002] The power fiber optic communication network is an important part of the national power grid and the power communication network. It is responsible for the functions of realizing real-time communication, remote control management and protection of the power system, and is a key infrastructure to ensure the normal operation, safety and stability of the power system. In recent years, due to the influence of extreme weather, natural disasters such as icing have continued to intensify the damage to transmission lines and power optical cables. How to accurately monitor the power grid operating environment in real time is a crucial research topic. Since the power fiber optic network structure is strongly coupled with the power grid structure and is in the same geographical space as the power grid, it provides a natural advantage for using power optical fiber to measure the environmental parameters of the power grid line. Using power optical fiber and sensing technology to monitor the status of the overhead optical cable of the power grid online, so as to objectively reflect the environmental parameters of the power line where the optical cable is located, has become a research trend and hot spot in recent years.

[0003] The integration of communication and perception is a new technology that has emerged in recent years. Although it was originally proposed in the field of wireless communications, it can also bring many benefits to long-distance optical fiber transmission systems. Based on the integration of interawareness, a new mode of hybrid transmission and mutual utilization of communication signals and sensor signals can be realized, which promotes the integrated development of sensor equipment and communication equipment. It is the best technical path to improve the distance and accuracy of sensor measurement, save network fiber core resources, and reduce operation and maintenance costs. It is the core issue of using power optical fiber to carry out environmental monitoring of overhead power lines, and has important research significance.

[0004] At present, there are three main integrated synaesthesia schemes proposed for long-distance optical fiber transmission systems: wavelength division multiplexing, space division multiplexing, and fully integrated waveforms. Due to the simple implementation method and the ability to share spectrum, the scheme based on wavelength division multiplexing has been widely used in commercial synaesthesia fusion systems of power grids. The peak power of the signal pulse of distributed optical fiber sensing is very high. If the communication signal is transmitted in the same direction as it, it will be seriously affected by the nonlinearity of the optical fiber. Therefore, in the optical fiber synaesthesia integration system based on wavelength division multiplexing, the communication signal and the high-power pulse signal are propagated in the opposite direction, so that the effective interaction time / length between the data channel and the sensing pulse signal is extremely short, which greatly reduces the signal interference caused by the nonlinearity of the optical fiber. The backscattered sensing signal propagates in the same direction as the data channel, but because the optical power is very weak, it will not affect the performance of the data channel. The synaesthesia integrated signal is sent to the coherent receiver through the wavelength division multiplexer for the communication channel and distributed optical fiber sensing.

[0005] In order to increase the transmission and perception distance, the integrated fiber synaesthesia system based on wavelength division multiplexing can use distributed Raman amplifiers for optical power compensation and control. However, unlike the traditional wavelength division multiplexing communication system, the integrated synaesthesia signal includes both communication signals and sensing signals. The communication signal pays more attention to the communication capacity, while the sensing signal pays more attention to indicators such as signal-to-noise ratio and sensitivity. Although the power of the reverse Rayleigh scattering sensing signal is weak and will not affect the communication signal, the nonlinear interference generated by the communication signal will be superimposed on the Rayleigh backscattering signal, affecting the signal-to-noise ratio of the sensing process. Furthermore, for the integrated fiber synaesthesia system that can perceive multi-dimensional information, there are multiple perception channels, and the functions of different perception channels are different, and the requirements for key indicators such as signal-to-noise ratio and resolution are also different. At present, the parameter determination method for distributed Raman fiber amplifiers usually cannot meet the different needs of communication and sensing at the same time, and it is difficult to maximize the synaesthesia fusion gain. Summary of the invention

[0006] The present invention provides a parameter determination method, device, equipment and medium for a distributed Raman fiber amplifier. By determining and solving the optimization objective function under different target application scenarios of an optical fiber synaesthesia integrated system, the optimal link end power configuration of each wavelength channel is obtained, and then the target pump power of each pump source in the distributed Raman fiber amplifier is obtained based on a machine learning algorithm, which can simultaneously meet the different needs of communication and sensing and maximize the synaesthesia fusion gain.

[0007] In order to solve the above technical problems, a first aspect of an embodiment of the present invention provides a method for determining parameters of a distributed Raman fiber amplifier, comprising:

[0008] Based on the target application scenario of the fiber optic interaceptive integrated system, determine the optimization objective function of the link end power configuration of each wavelength channel;

[0009] Solving the optimization objective function using a simulated annealing algorithm to obtain an optimal link end power configuration for each of the wavelength channels;

[0010] According to the fiber input optical power of each wavelength channel and the optimal link end power configuration, a pre-trained neural network model is used to obtain the target pump power of each pump source in the distributed Raman fiber amplifier.

[0011] As a preferred solution, the target application scenario of the optical fiber synaesthesia integrated system is based on which the optimization objective function of the link end power configuration of each wavelength channel is determined, specifically including:

[0012] When the target application scenario is a communication function priority scenario, the optimization objective is to maximize the capacity of the communication channel and the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the communication function priority scenario;

[0013] When the target application scenario is a sensing function priority scenario, the optimization goal is to maximize the effective signal-to-noise ratio of the sensing channel and the capacity of the communication channel meets the preset minimum capacity requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the sensing function priority scenario.

[0014] As a preferred solution, the optimization objective function is specifically:

[0015]

[0016] Where y represents the optimization objective function; n represents the number of communication channels; C i represents the capacity of the i-th communication channel, m represents the number of sensing channels; SNR eff,sj represents the effective signal-to-noise ratio of the jth sensing channel; w1 is used to represent the target application scenario. When w1=1, the target application scenario is the communication function priority scenario; when w1=0, the target application scenario is the sensing function priority scenario.

[0017] As a preferred solution, the use of a simulated annealing algorithm to solve the optimization objective function to obtain the optimal link end power configuration of each wavelength channel specifically includes:

[0018] Solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier;

[0019] Determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length;

[0020] Determine the optimized effective signal-to-noise ratio of each of the wavelength channels according to the Raman gain, the optical fiber input power of each of the wavelength channels, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient;

[0021] The optimal link end power configuration of each of the wavelength channels is determined according to the Raman gain and the optimized effective signal-to-noise ratio.

[0022] As a preferred solution, the method of solving the optimization objective function by using a simulated annealing algorithm to obtain the initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier specifically includes:

[0023] According to the preset maximum temperature, pump power search space and pump wavelength search space, the temperature, pump power and pump wavelength are randomly initialized;

[0024] Calculate the updated value of the current pump power and the updated value of the pump wavelength;

[0025] Determining an updated value of a current optimization objective function value according to the optimization objective function, an updated value of a current pump power, and an updated value of a pump wavelength;

[0026] According to a comparison result between an updated value of a current optimization objective function value and the current optimization objective function value, the current pump power, the pump wavelength, the optimal pump power, the optimal pump wavelength and the optimal optimization objective function value are updated;

[0027] Add 1 to the current number of iterations, and update the current temperature based on a preset cooling strategy, and add the current optimal pump power, optimal pump wavelength, and optimal optimization objective function value to the pump power optimization process function, the pump wavelength optimization process function, and the optimization objective function value optimization process function, respectively; re-execute the steps of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of iterations reaches a preset iteration number threshold; wherein the initial value of the number of iterations is 0;

[0028] Optimizing the process function based on the current optimization objective function value, updating the current number of retention times; resetting the current number of iterations, and re-executing the step of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature; wherein the initial value of the number of retention times is 0;

[0029] The initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier are obtained according to the optimal pump power added in the current pump power optimization process function and the optimal pump wavelength added in the pump wavelength optimization process function.

[0030] As a preferred solution, the step of calculating the updated value of the current pump power and the updated value of the pump wavelength specifically includes:

[0031] The updated value of the current pump power and the updated value of the pump wavelength are calculated by the following expressions:

[0032]

[0033] Among them, x new Indicates the updated value of the current pump power or pump wavelength; x current represents the current pump power or pump wavelength; r represents a random number between -1 and 1; sign() is a sign function; T represents the current temperature; Up and Low are the maximum and minimum values ​​of the pump power search space or the pump wavelength search space, respectively.

[0034] As a preferred solution, the current pump power, pump wavelength, optimal pump power, optimal pump wavelength and optimal optimization objective function value are updated according to the comparison result between the updated value of the current optimization objective function value and the current optimization objective function value, specifically including:

[0035] When exp(-df / T)>rand(0,1), the current pump power is updated to the updated value of the current pump power, the current pump wavelength is updated to the updated value of the current pump wavelength, and the current optimization objective function value is updated to the updated value of the current optimization objective function value; wherein df is the difference between the updated value of the current optimization objective function value and the current optimization objective function value; T represents the current temperature; rand(0,1) represents a random number between 0 and 1;

[0036] When exp(-df / T)>rand(0,1) and df>0, the current optimal pump power is updated to the updated value of the current pump power, the current optimal pump wavelength is updated to the updated value of the current pump wavelength, and the current optimal optimization objective function value is updated to the updated value of the current optimization objective function value.

[0037] As a preferred solution, the updating of the current temperature based on a preset cooling strategy specifically includes:

[0038] Update the current temperature using the following expression:

[0039] T=T max exp[-exp(-1) epoch] ;

[0040] Where, T represents the current temperature; T max Indicates the maximum temperature; epoch indicates the current iteration number.

[0041] As a preferred solution, the optimizing process function based on the current optimization objective function value and updating the current number of retention times specifically include:

[0042] When the difference between the last two items in the optimization process function of the current optimization objective function value is less than 0.001, the current retention times are increased by 1;

[0043] When the difference between the last two items in the current optimization objective function value optimization process function is greater than or equal to 0.001, the current retention times are set to 0.

[0044] As a preferred solution, the determining of the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length specifically includes:

[0045] When the distributed Raman fiber amplifier adopts a forward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0046]

[0047] When the distributed Raman fiber amplifier adopts a backward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0048]

[0049] Among them, G f,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts the forward pumping configuration; G b,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts a backward pumping configuration; t represents the number of pump sources; g R represents the Raman gain coefficient; ω i represents the central angular frequency of the i-th pump source; ω j represents the jth pump source central angular frequency; the pump source central angular frequency is calculated based on the initial optimized pump wavelength and the light propagation speed; A eff represents the effective cross-sectional area of ​​the optical fiber; α represents the attenuation coefficient of the optical fiber; L represents the physical length of the optical fiber; z represents the effective length of the optical fiber; P pj (0) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the forward pump configuration; P pj (L) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the backward pumping configuration.

[0050] As a preferred solution, determining the optimized effective signal-to-noise ratio of each wavelength channel according to the Raman gain, the optical fiber input power of each wavelength channel, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient specifically includes:

[0051] According to the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient, the cross-phase modulation nonlinear noise coefficient and the fiber-input optical power of each wavelength channel, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel are respectively determined by the following expressions:

[0052]

[0053]

[0054] According to the Raman gain, the optical fiber input power of each of the wavelength channels, the total power loss, the ASE noise, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel, the optimized effective signal-to-noise ratio of each of the wavelength channels is determined by the following expression:

[0055]

[0056] Among them, C ∞ represents the correlation coefficient of the self-channel nonlinear noise; η SPM,j represents the self-phase modulation nonlinear noise coefficient; η XPM,ij represents the nonlinear noise coefficient of the cross-phase modulation; P NLI,ci represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel; P ci (0) represents the optical power of the fiber input of the i-th communication channel; P cj (0) represents the optical power of the jth communication channel; P NLI,si represents the nonlinear crosstalk power at the end of the optical fiber of the i-th sensing channel; P si (0) represents the optical power of the i-th sensing channel; SNR link,i represents the signal-to-noise ratio of the i-th wavelength channel; G i A represents the Raman gain of the i-th wavelength channel at each position; i represents the total power loss of the i-th wavelength channel; P i (0) represents the optical power of the i-th wavelength channel; P ASE represents the ASE noise; P NLI,irepresents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel or the sensing channel; SNR eff,i represents the effective signal-to-noise ratio of the i-th wavelength channel; δ and β are both constants.

[0057] As a preferred solution, the method specifically obtains the neural network model through training through the following steps:

[0058] Based on the initial optimized pump power configuration and the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm, respectively determining the pump power scanning range and the fixed pump wavelength of each of the pump sources;

[0059] Based on the pump power scanning range and the fixed pump wavelength, using a spectrometer to measure the output optical power of each of the wavelength channels under different fiber input optical powers and pump powers;

[0060] Constructing a training data set using the input fiber optical power, the pump power and the output optical power;

[0061] Inputting the training data set into a deep neural network, training the deep neural network with minimizing the loss function as the optimization goal, and obtaining the neural network model;

[0062] The loss function is specifically:

[0063]

[0064] t represents the number of pump sources; represents the actual measured pump power of the i-th pump source; represents the pump power of the i-th pump source output by the deep neural network during each round of training.

[0065] A second aspect of an embodiment of the present invention provides a parameter determination device for a distributed Raman fiber amplifier, comprising:

[0066] An optimization objective function determination module is used to determine the optimization objective function of the link end power configuration of each wavelength channel based on the target application scenario of the optical fiber interawareness integrated system;

[0067] An optimization objective function solving module, used to solve the optimization objective function using a simulated annealing algorithm to obtain an optimal link end power configuration for each of the wavelength channels;

[0068] The target pump power determination module is used to obtain the target pump power of each pump source in the distributed Raman fiber amplifier using a pre-trained neural network model according to the fiber input optical power of each wavelength channel and the optimal link end power configuration.

[0069] As a preferred solution, the optimization objective function determination module is used to determine the optimization objective function of the link end power configuration of each wavelength channel based on the target application scenario of the optical fiber synaesthesia integrated system, specifically including:

[0070] When the target application scenario is a communication function priority scenario, the optimization objective is to maximize the capacity of the communication channel and the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the communication function priority scenario;

[0071] When the target application scenario is a sensing function priority scenario, the optimization goal is to maximize the effective signal-to-noise ratio of the sensing channel and the capacity of the communication channel meets the preset minimum capacity requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the sensing function priority scenario.

[0072] As a preferred solution, the optimization objective function is specifically:

[0073]

[0074] Where y represents the optimization objective function; n represents the number of communication channels; C i represents the capacity of the i-th communication channel, m represents the number of sensing channels; SNR eff,sj represents the effective signal-to-noise ratio of the jth sensing channel; w1 is used to represent the target application scenario. When w1=1, the target application scenario is the communication function priority scenario; when w1=0, the target application scenario is the sensing function priority scenario.

[0075] As a preferred solution, the optimization objective function solving module is used to solve the optimization objective function using a simulated annealing algorithm to obtain the optimal link end power configuration of each wavelength channel, specifically including:

[0076] Solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier;

[0077] Determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length;

[0078] Determine the optimized effective signal-to-noise ratio of each of the wavelength channels according to the Raman gain, the optical fiber input power of each of the wavelength channels, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient;

[0079] The optimal link end power configuration of each of the wavelength channels is determined according to the Raman gain and the optimized effective signal-to-noise ratio.

[0080] As a preferred solution, the optimization objective function solving module is used to solve the optimization objective function using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier, specifically including:

[0081] According to the preset maximum temperature, pump power search space and pump wavelength search space, the temperature, pump power and pump wavelength are randomly initialized;

[0082] Calculate the updated value of the current pump power and the updated value of the pump wavelength;

[0083] Determining an updated value of a current optimization objective function value according to the optimization objective function, an updated value of a current pump power, and an updated value of a pump wavelength;

[0084] According to a comparison result between an updated value of a current optimization objective function value and the current optimization objective function value, the current pump power, the pump wavelength, the optimal pump power, the optimal pump wavelength and the optimal optimization objective function value are updated;

[0085] Add 1 to the current number of iterations, and update the current temperature based on a preset cooling strategy, and add the current optimal pump power, optimal pump wavelength, and optimal optimization objective function value to the pump power optimization process function, the pump wavelength optimization process function, and the optimization objective function value optimization process function, respectively; re-execute the steps of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of iterations reaches a preset iteration number threshold; wherein the initial value of the number of iterations is 0;

[0086] Optimizing the process function based on the current optimization objective function value, updating the current number of retention times; resetting the current number of iterations, and re-executing the step of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature; wherein the initial value of the number of retention times is 0;

[0087] The initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier are obtained according to the optimal pump power added in the current pump power optimization process function and the optimal pump wavelength added in the pump wavelength optimization process function.

[0088] As a preferred solution, the optimization objective function solving module is used to calculate the updated value of the current pump power and the updated value of the pump wavelength, specifically including:

[0089] The updated value of the current pump power and the updated value of the pump wavelength are calculated by the following expressions:

[0090]

[0091] Among them, x new Indicates the updated value of the current pump power or pump wavelength; x current represents the current pump power or pump wavelength; r represents a random number between -1 and 1; sign() is a sign function; T represents the current temperature; Up and Low are the maximum and minimum values ​​of the pump power search space or the pump wavelength search space, respectively.

[0092] As a preferred solution, the optimization objective function solving module is used to update the current pump power, pump wavelength, optimal pump power, optimal pump wavelength and optimal optimization objective function value according to the comparison result between the updated value of the current optimization objective function value and the current optimization objective function value, specifically including:

[0093] When exp(-df / T)>rand(0,1), the current pump power is updated to the updated value of the current pump power, the current pump wavelength is updated to the updated value of the current pump wavelength, and the current optimization objective function value is updated to the updated value of the current optimization objective function value; wherein df is the difference between the updated value of the current optimization objective function value and the current optimization objective function value; T represents the current temperature; rand(0,1) represents a random number between 0 and 1;

[0094] When exp(-df / T)>rand(0,1) and df>0, the current optimal pump power is updated to the updated value of the current pump power, the current optimal pump wavelength is updated to the updated value of the current pump wavelength, and the current optimal optimization objective function value is updated to the updated value of the current optimization objective function value.

[0095] As a preferred solution, the optimization objective function solving module is used to update the current temperature based on a preset cooling strategy, specifically including:

[0096] Update the current temperature using the following expression:

[0097] T=T max exp[-exp(-1) epoch] ;

[0098] Where, T represents the current temperature; T max Indicates the maximum temperature; epoch indicates the current iteration number.

[0099] As a preferred solution, the optimization objective function solving module is used to optimize the process function based on the current optimization objective function value and update the current number of retention times, specifically including:

[0100] When the difference between the last two items in the optimization process function of the current optimization objective function value is less than 0.001, the current retention times are increased by 1;

[0101] When the difference between the last two items in the current optimization objective function value optimization process function is greater than or equal to 0.001, the current retention times are set to 0.

[0102] As a preferred solution, the optimization objective function solving module is used to determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length, specifically including:

[0103] When the distributed Raman fiber amplifier adopts a forward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0104]

[0105] When the distributed Raman fiber amplifier adopts a backward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0106]

[0107] Among them, G f,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts the forward pumping configuration; G b,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts a backward pumping configuration; t represents the number of pump sources; g R represents the Raman gain coefficient; ω i represents the central angular frequency of the i-th pump source; ω jrepresents the jth pump source central angular frequency; the pump source central angular frequency is calculated based on the initial optimized pump wavelength and the light propagation speed; A eff represents the effective cross-sectional area of ​​the optical fiber; α represents the attenuation coefficient of the optical fiber; L represents the physical length of the optical fiber; z represents the effective length of the optical fiber; P pj (0) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the forward pump configuration; P pj (L) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the backward pumping configuration.

[0108] As a preferred solution, the optimization objective function solving module is used to determine the optimized effective signal-to-noise ratio of each wavelength channel according to the Raman gain, the fiber optical power of each wavelength channel, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient, specifically including:

[0109] According to the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient, the cross-phase modulation nonlinear noise coefficient and the fiber-input optical power of each wavelength channel, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel are respectively determined by the following expressions:

[0110]

[0111] According to the Raman gain, the optical fiber input power of each of the wavelength channels, the total power loss, the ASE noise, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel, the optimized effective signal-to-noise ratio of each of the wavelength channels is determined by the following expression:

[0112]

[0113] Among them, C ∞ represents the correlation coefficient of the self-channel nonlinear noise; η SPM,j represents the self-phase modulation nonlinear noise coefficient; η XPM,ij represents the nonlinear noise coefficient of the cross-phase modulation; P NLI,ci represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel; P ci (0) represents the optical power of the fiber input of the i-th communication channel; P cj (0) represents the optical power of the jth communication channel; P NLI,si represents the nonlinear crosstalk power at the end of the optical fiber of the i-th sensing channel; P si(0) represents the optical power of the i-th sensing channel; SNR link,i represents the signal-to-noise ratio of the i-th wavelength channel; G i A represents the Raman gain of the i-th wavelength channel at each position; i represents the total power loss of the i-th wavelength channel; P i (0) represents the optical power of the i-th wavelength channel; P ASE represents the ASE noise; P NLI,i represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel or the sensing channel; SNR eff,i represents the effective signal-to-noise ratio of the i-th wavelength channel; δ and β are both constants.

[0114] As a preferred solution, the device further includes a neural network model training module, which is used to:

[0115] Based on the initial optimized pump power configuration and the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm, respectively determining the pump power scanning range and the fixed pump wavelength of each of the pump sources;

[0116] Based on the pump power scanning range and the fixed pump wavelength, using a spectrometer to measure the output optical power of each of the wavelength channels under different fiber input optical powers and pump powers;

[0117] Constructing a training data set using the input fiber optical power, the pump power and the output optical power;

[0118] Inputting the training data set into a deep neural network, training the deep neural network with minimizing the loss function as the optimization goal, and obtaining the neural network model;

[0119] The loss function is specifically:

[0120]

[0121] t represents the number of pump sources; represents the actual measured pump power of the i-th pump source; represents the pump power of the i-th pump source output by the deep neural network during each round of training.

[0122] A third aspect of an embodiment of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for determining parameters of a distributed Raman fiber amplifier as described in any one of the first aspects is implemented.

[0123] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the parameter determination method of the distributed Raman fiber amplifier as described in any one of the first aspects.

[0124] Compared with the prior art, the beneficial effect of the embodiments of the present invention lies in that by determining and solving the optimization objective function under different target application scenarios of the optical fiber synaesthesia integrated system, the optimal link end power configuration of each wavelength channel is obtained, and then the target pump power of each pump source in the distributed Raman fiber amplifier is obtained based on the machine learning algorithm, which can simultaneously meet the different needs of communication and sensing and maximize the synaesthesia fusion gain. BRIEF DESCRIPTION OF THE DRAWINGS

[0125] Figure 1 1 is a flow chart of a method for determining parameters of a distributed Raman fiber amplifier in an embodiment of the present invention;

[0126] Figure 2 is an architecture diagram of an experimental measurement system for a training data set of a deep neural network in an embodiment of the present invention;

[0127] Figure 3 is a schematic diagram of the architecture of a deep neural network in an embodiment of the present invention;

[0128] Figure 4 It is a schematic diagram of the structure of a parameter determination device for a distributed Raman fiber amplifier in an embodiment of the present invention. DETAILED DESCRIPTION

[0129] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0130] See also Figure 1 According to a first aspect of an embodiment of the present invention, a method for determining parameters of a distributed Raman fiber amplifier is provided, comprising the following steps S1 to S3:

[0131] Step S1, based on the target application scenario of the optical fiber interawareness integrated system, determining the optimization objective function of the link end power configuration of each wavelength channel;

[0132] Step S2, solving the optimization objective function using a simulated annealing algorithm to obtain the optimal link end power configuration of each of the wavelength channels;

[0133] Step S3, according to the fiber input optical power of each wavelength channel and the optimal link end power configuration, using a pre-trained neural network model to obtain the target pump power of each pump source in the distributed Raman fiber amplifier.

[0134] It is worth noting that in the integrated fiber synaesthesia system, in order to reduce the impact of high-power pulse sensing signals on the communication channel, the communication signal and the pulse sensing signal propagate in opposite directions, and the communication signal and the Rayleigh backscattering signal are transmitted in the same direction. A distributed Raman fiber amplifier is used to compensate and control the power, and the pump configuration can be forward, backward or bidirectional. The wavelength division multiplexing synaesthesia integrated system integrates multiple communication channels and sensing channels in the C band based on wavelength division multiplexing. The frequency interval between wavelength channels is fixed. In order to realize multi-dimensional information perception, multiple sensing wavelength channels are reserved in advance.

[0135] In this embodiment, different optimization objective functions are determined according to different target application scenarios and functions of the optical fiber synaesthesia integrated system. The performance index of each communication channel is set as the channel capacity C i =B·log2(1+SNR eff,i ), B is the channel bandwidth, and the total communication capacity is C = ΣC i The performance index of each sensing channel is set as the effective signal-to-noise ratio SNR eff,s . For the communication channel, the higher the effective signal-to-noise ratio, the larger the channel capacity and the larger the channel power, but the more serious the nonlinear noise generated in the sensing channel. For the sensing channel, the higher the effective signal-to-noise ratio, the higher the sensitivity and resolution. At this time, the nonlinear noise generated by the communication channel is required to be smaller, that is, the power of the communication channel is required to be reduced. Therefore, there is a balance between the capacity of the communication channel and the signal-to-noise ratio of the sensing channel. This embodiment determines the optimization objective function of the link end power configuration of each wavelength channel according to whether the target application scenario is a communication function priority scenario or a sensing function priority scenario, so as to simultaneously meet the different needs of communication and sensing and maximize the synaesthesia fusion gain.

[0136] Furthermore, this embodiment uses a simulated annealing algorithm to solve the optimization objective function. The basic idea of ​​the simulated annealing algorithm is that during the search process, the algorithm accepts a poor solution with a certain probability, so that it can jump out of the local optimal solution and explore a wider search space. Finally, the optimal link end power configuration of each wavelength channel that can achieve the optimization goal is obtained.

[0137] Furthermore, in order to improve the design accuracy of the pump power of each pump source in the distributed Raman fiber amplifier, this embodiment uses a pre-trained neural network model based on the input fiber optical power of each wavelength channel and the optimal link end power configuration to output the target pump power of each pump source.

[0138] As a preferred solution, the target application scenario of the optical fiber synaesthesia integrated system is based on which the optimization objective function of the link end power configuration of each wavelength channel is determined, specifically including:

[0139] When the target application scenario is a communication function priority scenario, the optimization objective is to maximize the capacity of the communication channel and the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the communication function priority scenario;

[0140] When the target application scenario is a sensing function priority scenario, the optimization goal is to maximize the effective signal-to-noise ratio of the sensing channel and the capacity of the communication channel meets the preset minimum capacity requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the sensing function priority scenario.

[0141] It is worth noting that, in this embodiment, if the target application scenario of the optical fiber synaesthesia integrated system is a communication function priority scenario, the following optimization goals are set:

[0142]

[0143] Where C = B·log2(1+SNR eff ), represents the communication channel capacity, which maximizes the capacity of the communication channel while ensuring that the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement.

[0144] If the target application scenario of the fiber optic interaceptive integrated system is a sensing function priority scenario, the following optimization goals are set:

[0145]

[0146] This ensures that the capacity of the communication channel meets the preset minimum capacity requirement while maximizing the effective signal-to-noise ratio of the sensing channel.

[0147] As a preferred solution, the optimization objective function is specifically:

[0148]

[0149] Where y represents the optimization objective function; n represents the number of communication channels; C i represents the capacity of the i-th communication channel, m represents the number of sensing channels; SNR eff,sj represents the effective signal-to-noise ratio of the jth sensing channel; w1 is used to represent the target application scenario. When w1=1, the target application scenario is the communication function priority scenario; when w1=0, the target application scenario is the sensing function priority scenario.

[0150] Specifically, in order to solve the balance optimization problem between the capacity of the communication channel and the signal-to-noise ratio of the sensing channel, this embodiment defines the above-mentioned optimization objective function. When w1=1, it corresponds to the target application scenario with full communication function priority. When w1=0, it corresponds to the target application scenario with full sensing function priority. In other cases, a balance is struck between the communication function and the sensing function.

[0151] As a preferred solution, the use of a simulated annealing algorithm to solve the optimization objective function to obtain the optimal link end power configuration of each wavelength channel specifically includes:

[0152] Solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier;

[0153] Determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length;

[0154] Determine the optimized effective signal-to-noise ratio of each of the wavelength channels according to the Raman gain, the optical fiber input power of each of the wavelength channels, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient;

[0155] The optimal link end power configuration of each of the wavelength channels is determined according to the Raman gain and the optimized effective signal-to-noise ratio.

[0156] Specifically, in order to effectively calculate the optimal link end power configuration of each wavelength channel, it is necessary to clarify the signal-to-noise ratio introduced by the synaesthesia signal integration and establish an effective signal-to-noise ratio model that can consider the transceiver execution cost and Raman gain.

[0157] When integrating communication and sensor signals, optical couplers, optical beam splitters and other devices are used, and the transceiver itself also has defects. The impact of these factors on the interaural integration system can be uniformly attributed to the additional signal-to-noise ratio introduced by the transceiver, so it needs to be accurately measured. Assuming that the additional signal-to-noise ratio introduced by the transceiver has nothing to do with the channel wavelength, the measurement method is as follows:

[0158] Step 1: SNR eff is the effective signal-to-noise ratio of a wavelength channel, SNR link It is the signal-to-noise ratio generated on the entire optical fiber link, mainly including ASE noise and nonlinear noise, so the calculation formula of the effective signal-to-noise ratio is as follows:

[0159]

[0160] Further rewritten as:

[0161]

[0162] Further rewritten as:

[0163]

[0164] Step 2: In a back-to-back (BtB) scenario, measure the bit error rate BER-optical signal-to-noise ratio OSNR data and draw a curve.

[0165] Step 3: For a given OSNR experimentally measured in a BtB scenario, use the formula: SNR = OSNR · Δv / B ch Convert it to signal-to-noise ratio. Since there is no nonlinear noise in the BtB scenario, the signal-to-noise ratio is the SNR at this time. link .

[0166] Step 4: Find the corresponding BER experimental value from the OSNR and substitute the BER into the formula:

[0167]

[0168] Inversely deduce a new signal-to-noise ratio, which is the corresponding SNR at this time eff .

[0169] Step 5: Based on a set of SNR link and SNR eff Data, draw a curve graph, the horizontal axis is 1 / SNR link , the vertical axis is 1 / SNR eff , the slope of the straight line is 1 / δ, and the intercept of the ordinate is β / δ, from which the values ​​of δ and β can be determined. This is the linear least squares fitting process.

[0170] Regarding the process of establishing an effective signal-to-noise ratio model, first of all, the entire fiber-optic intersensing integrated system is an ultra-long-distance system, which only contains one fiber span with a length of L. The optical signal that realizes the sensing function is a backscattered signal with a relatively weak optical power. Therefore, when establishing the communication channel power model, the nonlinear interference of the sensing signal on the communication signal is ignored. The optical power of the communication signal is relatively strong, and the sensing signal is relatively sensitive to the signal-to-noise ratio. Therefore, when establishing the sensing channel power model, the nonlinear interference of the communication signal on the sensing signal is considered, and the mutual influence between different sensing signals is ignored. The input fiber power of each communication signal is denoted as {P c1 (0), P c2 (0),…,P cn (0)}, the channel center frequency is denoted as {f c1 , fc2 , …, f cn}, the power at the end of the optical fiber link is denoted as {P c1 (L), P c2 (L),…,P cn (L)}; the fiber power of each sensor signal is recorded as {P s1 (0), P s2 (0),…,P sm (0)}, the channel center frequency is denoted as {f s1 , f s2 , …, f sm}, the power at the end of the optical fiber link is denoted as {P s1 (L), P s2 (L),…,P sm (L)}. The fiber Raman amplifier contains t pump sources, and the pump power is denoted as {P p1 (0), P p2 (0),…,P pt (0)}, the pump center frequency is denoted as {f p1 , f p2 , …, f pt}.

[0171] For each communication channel, a Gaussian noise model is used to calculate the nonlinear crosstalk power. In this embodiment, the nonlinear crosstalk power at the end of the optical fiber of the communication channel is estimated according to the following formula:

[0172]

[0173] For each sensing channel, since its own optical power is weak, ignoring the nonlinear interference induced by self-phase modulation and the mutual influence between different sensing channels, the nonlinear crosstalk power at the end of the optical fiber of the sensing channel is estimated according to the following formula:

[0174]

[0175] If the distributed Raman fiber amplifier adopts the forward pumping configuration, the gain of the distributed Raman fiber amplifier is calculated using the following formula:

[0176]

[0177] If the distributed Raman fiber amplifier adopts a backward pumping configuration, the gain of the distributed Raman fiber amplifier is calculated using the following formula:

[0178]

[0179] For each wavelength channel, the signal-to-noise ratio (SNR) generated on the entire fiber link link Defined as:

[0180]

[0181] In SNR link After the calculation is completed, substituting it into the above expression (4) can calculate the effective signal-to-noise ratio SNR of each channel eff .

[0182] Therefore, this embodiment first uses the simulated annealing algorithm to solve the above optimization objective function, and obtains the initial optimized pump power configuration and initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier corresponding to the optimal performance of the fiber optic interawareness integrated system. The Raman gain and optimized effective signal-to-noise ratio that can be achieved by each wavelength channel can be further calculated, and finally the optimal link end power configuration {P c1 (L), P c2 (L),…,P cn (L)} and {P s1 (L), P s2 (L),…,P sm (L)}.

[0183] As a preferred solution, the method of solving the optimization objective function by using a simulated annealing algorithm to obtain the initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier specifically includes:

[0184] According to the preset maximum temperature, pump power search space and pump wavelength search space, the temperature, pump power and pump wavelength are randomly initialized;

[0185] Calculate the updated value of the current pump power and the updated value of the pump wavelength;

[0186] Determining an updated value of a current optimization objective function value according to the optimization objective function, an updated value of a current pump power, and an updated value of a pump wavelength;

[0187] According to a comparison result between an updated value of a current optimization objective function value and the current optimization objective function value, the current pump power, the pump wavelength, the optimal pump power, the optimal pump wavelength and the optimal optimization objective function value are updated;

[0188] Add 1 to the current number of iterations, and update the current temperature based on a preset cooling strategy, and add the current optimal pump power, optimal pump wavelength, and optimal optimization objective function value to the pump power optimization process function, the pump wavelength optimization process function, and the optimization objective function value optimization process function, respectively; re-execute the steps of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of iterations reaches a preset iteration number threshold; wherein the initial value of the number of iterations is 0;

[0189] Optimizing the process function based on the current optimization objective function value, updating the current number of retention times; resetting the current number of iterations, and re-executing the step of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature; wherein the initial value of the number of retention times is 0;

[0190] The initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier are obtained according to the optimal pump power added in the current pump power optimization process function and the optimal pump wavelength added in the pump wavelength optimization process function.

[0191] Specifically, the present embodiment first needs to set a pump power search space and a pump wavelength search space, wherein the maximum value and the minimum value are respectively recorded as Up and Low, set the maximum temperature and the minimum temperature of the simulated annealing algorithm, and set the number of iterations at each temperature and the maximum number of retention times of the algorithm. In the initial state, the number of iterations and the number of retention times are both 0, and the initial value of the temperature is the maximum temperature set above. The pump power and the pump wavelength are randomly initialized in the pump power search space and the pump wavelength search space, and the simulated annealing algorithm is continuously optimized to obtain a global optimal solution. If the termination condition is met, that is, the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature, the algorithm is terminated to obtain the final initial optimized pump power configuration and the initial optimized pump wavelength configuration.

[0192] As a preferred solution, the step of calculating the updated value of the current pump power and the updated value of the pump wavelength specifically includes:

[0193] The updated value of the current pump power and the updated value of the pump wavelength are calculated by the following expressions:

[0194]

[0195] Among them, x new Indicates the updated value of the current pump power or pump wavelength; x current represents the current pump power or pump wavelength; r represents a random number between -1 and 1; sign() is a sign function; T represents the current temperature; Up and Low are the maximum and minimum values ​​of the pump power search space or the pump wavelength search space, respectively.

[0196] As a preferred solution, the current pump power, pump wavelength, optimal pump power, optimal pump wavelength and optimal optimization objective function value are updated according to the comparison result between the updated value of the current optimization objective function value and the current optimization objective function value, specifically including:

[0197] When exp(-df / T)>rand(0,1), the current pump power is updated to the updated value of the current pump power, the current pump wavelength is updated to the updated value of the current pump wavelength, and the current optimization objective function value is updated to the updated value of the current optimization objective function value; wherein df is the difference between the updated value of the current optimization objective function value and the current optimization objective function value; T represents the current temperature; rand(0,1) represents a random number between 0 and 1;

[0198] When exp(-df / T)>rand(0,1) and df>0, the current optimal pump power is updated to the updated value of the current pump power, the current optimal pump wavelength is updated to the updated value of the current pump wavelength, and the current optimal optimization objective function value is updated to the updated value of the current optimization objective function value.

[0199] As a preferred solution, the updating of the current temperature based on a preset cooling strategy specifically includes:

[0200] Update the current temperature using the following expression:

[0201] T=T max exp[-exp(-1) epoch] ;

[0202] Where, T represents the current temperature; T max Indicates the maximum temperature; epoch indicates the current iteration number.

[0203] As a preferred solution, the optimizing process function based on the current optimization objective function value and updating the current number of retention times specifically include:

[0204] When the difference between the last two items in the optimization process function of the current optimization objective function value is less than 0.001, the current retention times are increased by 1;

[0205] When the difference between the last two items in the current optimization objective function value optimization process function is greater than or equal to 0.001, the current retention times are set to 0.

[0206] As a preferred solution, the determining of the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length specifically includes:

[0207] When the distributed Raman fiber amplifier adopts a forward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0208]

[0209] When the distributed Raman fiber amplifier adopts a backward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0210]

[0211] Among them, G f,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts the forward pumping configuration; G b,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts a backward pumping configuration; t represents the number of pump sources; g R represents the Raman gain coefficient; ω i represents the central angular frequency of the i-th pump source; ω j represents the jth pump source central angular frequency; the pump source central angular frequency is calculated based on the initial optimized pump wavelength and the light propagation speed; A eff represents the effective cross-sectional area of ​​the optical fiber; α represents the attenuation coefficient of the optical fiber; L represents the physical length of the optical fiber; z represents the effective length of the optical fiber; P pj (0) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the forward pump configuration; P pj (L) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the backward pumping configuration.

[0212] As a preferred solution, determining the optimized effective signal-to-noise ratio of each wavelength channel according to the Raman gain, the optical fiber input power of each wavelength channel, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient specifically includes:

[0213] According to the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient, the cross-phase modulation nonlinear noise coefficient and the fiber-input optical power of each wavelength channel, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel are respectively determined by the following expressions:

[0214]

[0215] According to the Raman gain, the optical fiber input power of each of the wavelength channels, the total power loss, the ASE noise, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel, the optimized effective signal-to-noise ratio of each of the wavelength channels is determined by the following expression:

[0216]

[0217] Among them, C ∞ represents the correlation coefficient of the self-channel nonlinear noise; η SPM,j represents the self-phase modulation nonlinear noise coefficient; η XPM,ij represents the nonlinear noise coefficient of the cross-phase modulation; P NLI,ci represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel; P ci (0) represents the optical power of the fiber input of the i-th communication channel; P cj (0) represents the optical power of the jth communication channel; P NLI,si represents the nonlinear crosstalk power at the end of the optical fiber of the i-th sensing channel; P si (0) represents the optical power of the i-th sensing channel; SNR link,i represents the signal-to-noise ratio of the i-th wavelength channel; G i A represents the Raman gain of the i-th wavelength channel at each position; i represents the total power loss of the i-th wavelength channel; P i (0) represents the optical power of the i-th wavelength channel; P ASE represents the ASE noise; P NLI,i represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel or the sensing channel; SNR eff,i represents the effective signal-to-noise ratio of the i-th wavelength channel; δ and β are both constants.

[0218] As a preferred solution, the method specifically obtains the neural network model through training through the following steps:

[0219] Based on the initial optimized pump power configuration and the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm, respectively determining the pump power scanning range and the fixed pump wavelength of each of the pump sources;

[0220] Based on the pump power scanning range and the fixed pump wavelength, using a spectrometer to measure the output optical power of each of the wavelength channels under different fiber input optical powers and pump powers;

[0221] Constructing a training data set using the input fiber optical power, the pump power and the output optical power;

[0222] Inputting the training data set into a deep neural network, training the deep neural network with minimizing the loss function as the optimization goal, and obtaining the neural network model;

[0223] The loss function is specifically:

[0224]

[0225] t represents the number of pump sources; represents the actual measured pump power of the i-th pump source; represents the pump power of the i-th pump source output by the deep neural network during each round of training.

[0226] Specifically, the advantage of the above effective signal-to-noise ratio model is that it can be calculated analytically and quickly. However, some parameters are estimated approximately when the model is established, so the calculation accuracy of the pump power and pump wavelength configuration is limited. In order to improve the design accuracy, the reverse design model of the distributed Raman fiber amplifier is established using experimental data and artificial neural networks. First, build Figure 2 The experimental system shown in the figure takes the wavelength of the pump source as a fixed value based on the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm; and determines the pump power scanning range of each pump source based on the initial optimized pump power configuration obtained by solving the optimization objective function using the simulated annealing algorithm. By changing the fiber input light power and the pump power, the pump power can be scanned at different fiber input light powers. N = n + m, and the pump power Under the condition of The above-mentioned input fiber optical power, pump power and output optical power are used to construct a training data set and establish Figure 3 The deep neural network shown in FIG. When training the deep neural network, the input data is the input optical power and the output optical power. With the optimization goal of minimizing the above loss function, each network parameter is optimized and trained to obtain a trained neural network model. In practical applications, it is only necessary to configure the input optical power of each wavelength channel and the optimal link end power that can achieve the optimization goal to output the corresponding pump power of each pump source.

[0227] The parameter determination method of the distributed Raman fiber amplifier provided in the embodiment of the present invention obtains the optimal link end power configuration of each wavelength channel by determining and solving the optimization objective function under different target application scenarios of the optical fiber synaesthesia integrated system, and then obtains the target pump power of each pump source in the distributed Raman fiber amplifier based on the machine learning algorithm, which can simultaneously meet the different needs of communication and sensing and maximize the synaesthesia fusion gain.

[0228] See also Figure 4 According to a second aspect of an embodiment of the present invention, there is provided a device for determining parameters of a distributed Raman fiber amplifier, comprising:

[0229] The optimization objective function determination module 11 is used to determine the optimization objective function of the link end power configuration of each wavelength channel based on the target application scenario of the optical fiber interawareness integrated system;

[0230] An optimization objective function solving module 12 is used to solve the optimization objective function using a simulated annealing algorithm to obtain an optimal link end power configuration for each of the wavelength channels;

[0231] The target pump power determination module 13 is used to obtain the target pump power of each pump source in the distributed Raman fiber amplifier using a pre-trained neural network model according to the fiber input optical power of each wavelength channel and the optimal link end power configuration.

[0232] As a preferred solution, the optimization objective function determination module 11 is used to determine the optimization objective function of the link end power configuration of each wavelength channel based on the target application scenario of the optical fiber synaesthesia integrated system, specifically including:

[0233] When the target application scenario is a communication function priority scenario, the optimization objective is to maximize the capacity of the communication channel and the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the communication function priority scenario;

[0234] When the target application scenario is a sensing function priority scenario, the optimization goal is to maximize the effective signal-to-noise ratio of the sensing channel and the capacity of the communication channel meets the preset minimum capacity requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the sensing function priority scenario.

[0235] As a preferred solution, the optimization objective function is specifically:

[0236]

[0237] Where y represents the optimization objective function; n represents the number of communication channels; C irepresents the capacity of the i-th communication channel, m represents the number of sensing channels; SNR eff,sj represents the effective signal-to-noise ratio of the jth sensing channel; w1 is used to represent the target application scenario. When w1=1, the target application scenario is the communication function priority scenario; when w1=0, the target application scenario is the sensing function priority scenario.

[0238] As a preferred solution, the optimization objective function solving module 12 is used to solve the optimization objective function using a simulated annealing algorithm to obtain the optimal link end power configuration of each wavelength channel, specifically including:

[0239] Solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier;

[0240] Determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length;

[0241] Determine the optimized effective signal-to-noise ratio of each of the wavelength channels according to the Raman gain, the optical fiber input power of each of the wavelength channels, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient;

[0242] The optimal link end power configuration of each of the wavelength channels is determined according to the Raman gain and the optimized effective signal-to-noise ratio.

[0243] As a preferred solution, the optimization objective function solving module 12 is used to solve the optimization objective function using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier, specifically including:

[0244] According to the preset maximum temperature, pump power search space and pump wavelength search space, the temperature, pump power and pump wavelength are randomly initialized;

[0245] Calculate the updated value of the current pump power and the updated value of the pump wavelength;

[0246] Determining an updated value of a current optimization objective function value according to the optimization objective function, an updated value of a current pump power, and an updated value of a pump wavelength;

[0247] According to a comparison result between an updated value of a current optimization objective function value and the current optimization objective function value, the current pump power, the pump wavelength, the optimal pump power, the optimal pump wavelength and the optimal optimization objective function value are updated;

[0248] Add 1 to the current number of iterations, and update the current temperature based on a preset cooling strategy, and add the current optimal pump power, optimal pump wavelength, and optimal optimization objective function value to the pump power optimization process function, the pump wavelength optimization process function, and the optimization objective function value optimization process function, respectively; re-execute the steps of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of iterations reaches a preset iteration number threshold; wherein the initial value of the number of iterations is 0;

[0249] Optimizing the process function based on the current optimization objective function value, updating the current number of retention times; resetting the current number of iterations, and re-executing the step of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature; wherein the initial value of the number of retention times is 0;

[0250] The initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier are obtained according to the optimal pump power added in the current pump power optimization process function and the optimal pump wavelength added in the pump wavelength optimization process function.

[0251] As a preferred solution, the optimization objective function solving module 12 is used to calculate the updated value of the current pump power and the updated value of the pump wavelength, specifically including:

[0252] The updated value of the current pump power and the updated value of the pump wavelength are calculated by the following expressions:

[0253]

[0254] Among them, x new Indicates the updated value of the current pump power or pump wavelength; x current represents the current pump power or pump wavelength; r represents a random number between -1 and 1; sign() is a sign function; T represents the current temperature; Up and Low are the maximum and minimum values ​​of the pump power search space or the pump wavelength search space, respectively.

[0255] As a preferred solution, the optimization objective function solving module 12 is used to update the current pump power, pump wavelength, optimal pump power, optimal pump wavelength and optimal optimization objective function value according to the comparison result between the updated value of the current optimization objective function value and the current optimization objective function value, specifically including:

[0256] When exp(-df / T)>rand(0,1), the current pump power is updated to the updated value of the current pump power, the current pump wavelength is updated to the updated value of the current pump wavelength, and the current optimization objective function value is updated to the updated value of the current optimization objective function value; wherein df is the difference between the updated value of the current optimization objective function value and the current optimization objective function value; T represents the current temperature; rand(0,1) represents a random number between 0 and 1;

[0257] When exp(-df / T)>rand(0,1) and df>0, the current optimal pump power is updated to the updated value of the current pump power, the current optimal pump wavelength is updated to the updated value of the current pump wavelength, and the current optimal optimization objective function value is updated to the updated value of the current optimization objective function value.

[0258] As a preferred solution, the optimization objective function solving module 12 is used to update the current temperature based on a preset cooling strategy, specifically including:

[0259] Update the current temperature using the following expression:

[0260] T=T max exp[-exp(-1) epoch] ;

[0261] Where, T represents the current temperature; T max Indicates the maximum temperature; epoch indicates the current iteration number.

[0262] As a preferred solution, the optimization objective function solving module 12 is used to optimize the process function based on the current optimization objective function value and update the current number of retention times, specifically including:

[0263] When the difference between the last two items in the optimization process function of the current optimization objective function value is less than 0.001, the current retention times are increased by 1;

[0264] When the difference between the last two items in the current optimization objective function value optimization process function is greater than or equal to 0.001, the current retention times are set to 0.

[0265] As a preferred solution, the optimization objective function solving module 12 is used to determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length, specifically including:

[0266] When the distributed Raman fiber amplifier adopts a forward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0267]

[0268] When the distributed Raman fiber amplifier adopts a backward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression:

[0269]

[0270] Among them, G f,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts the forward pumping configuration; G b,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts a backward pumping configuration; t represents the number of pump sources; g R represents the Raman gain coefficient; ω i represents the central angular frequency of the i-th pump source; ω j represents the jth pump source central angular frequency; the pump source central angular frequency is calculated based on the initial optimized pump wavelength and the light propagation speed; A eff represents the effective cross-sectional area of ​​the optical fiber; α represents the attenuation coefficient of the optical fiber; L represents the physical length of the optical fiber; z represents the effective length of the optical fiber; P pj (0) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the forward pump configuration; P pj (L) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the backward pumping configuration.

[0271] As a preferred solution, the optimization objective function solving module 12 is used to determine the optimized effective signal-to-noise ratio of each wavelength channel according to the Raman gain, the fiber optical power of each wavelength channel, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient, specifically including:

[0272] According to the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient, the cross-phase modulation nonlinear noise coefficient and the fiber-input optical power of each wavelength channel, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel are respectively determined by the following expressions:

[0273]

[0274] According to the Raman gain, the optical fiber input power of each of the wavelength channels, the total power loss, the ASE noise, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel, the optimized effective signal-to-noise ratio of each of the wavelength channels is determined by the following expression:

[0275]

[0276] Among them, C ∞ represents the correlation coefficient of the self-channel nonlinear noise; η SPM,j represents the self-phase modulation nonlinear noise coefficient; η XPM,ij represents the nonlinear noise coefficient of the cross-phase modulation; P NLI,ci represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel; P ci (0) represents the optical power of the fiber input of the i-th communication channel; P cj (0) represents the optical power of the jth communication channel; P NLI,si represents the nonlinear crosstalk power at the end of the optical fiber of the i-th sensing channel; P si (0) represents the optical power of the i-th sensing channel; SNR link,i represents the signal-to-noise ratio of the i-th wavelength channel; G i A represents the Raman gain of the i-th wavelength channel at each position; i represents the total power loss of the i-th wavelength channel; P i (0) represents the optical power of the i-th wavelength channel; P ASE represents the ASE noise; P NLI,i represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel or the sensing channel; SNR eff,i represents the effective signal-to-noise ratio of the i-th wavelength channel; δ and β are both constants.

[0277] As a preferred solution, the device further includes a neural network model training module, which is used to:

[0278] Based on the initial optimized pump power configuration and the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm, respectively determining the pump power scanning range and the fixed pump wavelength of each of the pump sources;

[0279] Based on the pump power scanning range and the fixed pump wavelength, using a spectrometer to measure the output optical power of each of the wavelength channels under different fiber input optical powers and pump powers;

[0280] Constructing a training data set using the input fiber optical power, the pump power and the output optical power;

[0281] Inputting the training data set into a deep neural network, training the deep neural network with minimizing the loss function as the optimization goal, and obtaining the neural network model;

[0282] The loss function is specifically:

[0283]

[0284] t represents the number of pump sources; represents the actual measured pump power of the i-th pump source; represents the pump power of the i-th pump source output by the deep neural network during each round of training.

[0285] The parameter determination device of the distributed Raman fiber amplifier provided in the embodiment of the present invention obtains the optimal link end power configuration of each wavelength channel by determining and solving the optimization objective function under different target application scenarios of the optical fiber synaesthesia integrated system, and then obtains the target pump power of each pump source in the distributed Raman fiber amplifier based on the machine learning algorithm, which can simultaneously meet the different needs of communication and sensing and maximize the synaesthesia fusion gain.

[0286] A third aspect of an embodiment of the present invention provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for determining parameters of a distributed Raman fiber amplifier as described in any embodiment of the first aspect is implemented.

[0287] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, which are used to describe the execution process of the computer program in the electronic device.

[0288] The electronic device may include, but is not limited to, a processor and a memory. Those skilled in the art will appreciate that the schematic diagram is merely an example of an electronic device and does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the diagram, or may combine certain components, or different components, for example, the electronic device may also include an input / output device, a network access device, a bus, etc.

[0289] The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the electronic device, and uses various interfaces and lines to connect various parts of the entire electronic device.

[0290] The memory can be used to store the computer program and / or module, and the processor realizes various functions of the electronic device by running or executing the computer program and / or module stored in the memory, and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area can store data created according to the use of the mobile phone (such as audio data, a phone book, etc.), etc. In addition, the memory can include a high-speed random access memory, and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0291] A fourth aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the parameter determination method of the distributed Raman fiber amplifier as described in any one of the first aspects.

[0292] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary hardware platform, and of course can also be implemented entirely by hardware. Based on such an understanding, all or part of the contribution of the technical solution of the present invention to the background technology can be embodied in the form of a software product, and the computer software product can be stored in a storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0293] The above is a preferred embodiment of the present invention. It should be pointed out that a person skilled in the art can make several improvements and modifications without departing from the principle of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A method for determining parameters of a distributed Raman fiber amplifier, characterized in that: include: Based on the target application scenario of the fiber optic interaceptive integrated system, determine the optimization objective function of the link end power configuration of each wavelength channel; Solving the optimization objective function using a simulated annealing algorithm to obtain an optimal link end power configuration for each of the wavelength channels; According to the fiber input optical power of each wavelength channel and the optimal link end power configuration, a pre-trained neural network model is used to obtain the target pump power of each pump source in the distributed Raman fiber amplifier.

2. The method for determining parameters of a distributed Raman fiber amplifier according to claim 1, wherein: The target application scenario based on the optical fiber interawareness integrated system, determining the optimization objective function of the link end power configuration of each wavelength channel specifically includes: When the target application scenario is a communication function priority scenario, the optimization objective is to maximize the capacity of the communication channel and the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the communication function priority scenario; When the target application scenario is a sensing function priority scenario, the optimization goal is to maximize the effective signal-to-noise ratio of the sensing channel and the capacity of the communication channel meets the preset minimum capacity requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the sensing function priority scenario.

3. The method for determining parameters of a distributed Raman fiber amplifier according to claim 2, wherein: The optimization objective function is specifically: Where y represents the optimization objective function; n represents the number of communication channels; C i represents the capacity of the i-th communication channel, m represents the number of sensing channels; SNR eff,sj represents the effective signal-to-noise ratio of the jth sensing channel; w1 is used to represent the target application scenario. When w1=1, the target application scenario is the communication function priority scenario; when w1=0, the target application scenario is the sensing function priority scenario.

4. The method for determining parameters of a distributed Raman fiber amplifier according to claim 1, wherein: The using of a simulated annealing algorithm to solve the optimization objective function to obtain the optimal link end power configuration of each wavelength channel specifically includes: Solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier; Determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length; Determine the optimized effective signal-to-noise ratio of each of the wavelength channels according to the Raman gain, the optical fiber input power of each of the wavelength channels, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient; The optimal link end power configuration of each of the wavelength channels is determined according to the Raman gain and the optimized effective signal-to-noise ratio.

5. The method for determining parameters of a distributed Raman fiber amplifier according to claim 4, wherein: The method of solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier specifically includes: According to the preset maximum temperature, pump power search space and pump wavelength search space, the temperature, pump power and pump wavelength are randomly initialized; Calculate the updated value of the current pump power and the updated value of the pump wavelength; Determining an updated value of a current optimization objective function value according to the optimization objective function, an updated value of a current pump power, and an updated value of a pump wavelength; According to a comparison result between an updated value of a current optimization objective function value and the current optimization objective function value, the current pump power, the pump wavelength, the optimal pump power, the optimal pump wavelength and the optimal optimization objective function value are updated; Add 1 to the current number of iterations, and update the current temperature based on a preset cooling strategy, and add the current optimal pump power, optimal pump wavelength, and optimal optimization objective function value to the pump power optimization process function, the pump wavelength optimization process function, and the optimization objective function value optimization process function, respectively; re-execute the steps of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of iterations reaches a preset iteration number threshold; wherein the initial value of the number of iterations is 0; Optimizing the process function based on the current optimization objective function value, updating the current number of retention times; resetting the current number of iterations, and re-executing the step of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature; wherein the initial value of the number of retention times is 0; The initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier are obtained according to the optimal pump power added in the current pump power optimization process function and the optimal pump wavelength added in the pump wavelength optimization process function.

6. The method for determining parameters of a distributed Raman fiber amplifier according to claim 5, characterized in that: The calculating of the updated value of the current pump power and the updated value of the pump wavelength specifically includes: The updated value of the current pump power and the updated value of the pump wavelength are calculated by the following expressions: Among them, x new Indicates the updated value of the current pump power or pump wavelength; x current represents the current pump power or pump wavelength; r represents a random number between -1 and 1; sign() is a sign function; T represents the current temperature; Up and Low are the maximum and minimum values ​​of the pump power search space or the pump wavelength search space, respectively.

7. The method for determining parameters of a distributed Raman fiber amplifier according to claim 5, characterized in that: The updating of the current pump power, pump wavelength, optimal pump power, optimal pump wavelength and optimal optimization objective function value according to the comparison result between the updated value of the current optimization objective function value and the current optimization objective function value specifically includes: When exp(-df / T)>rand(0,1), the current pump power is updated to the updated value of the current pump power, the current pump wavelength is updated to the updated value of the current pump wavelength, and the current optimization objective function value is updated to the updated value of the current optimization objective function value; wherein df is the difference between the updated value of the current optimization objective function value and the current optimization objective function value; T represents the current temperature; rand(0,1) represents a random number between 0 and 1; When exp(-df / T)>rand(0,1) and df>0, the current optimal pump power is updated to the updated value of the current pump power, the current optimal pump wavelength is updated to the updated value of the current pump wavelength, and the current optimal optimization objective function value is updated to the updated value of the current optimization objective function value.

8. The method for determining parameters of a distributed Raman fiber amplifier according to claim 5, wherein: The updating of the current temperature based on the preset cooling strategy specifically includes: Update the current temperature using the following expression: T=T max ·exp[-exp(-1)·epoch]。; Where, T represents the current temperature; T max Indicates the maximum temperature; epoch indicates the current iteration number.

9. The method for determining parameters of a distributed Raman fiber amplifier according to claim 5, wherein: The optimizing process function based on the current optimization objective function value and updating the current number of retention times specifically include: When the difference between the last two items in the optimization process function of the current optimization objective function value is less than 0.001, the current retention times are increased by 1; When the difference between the last two items in the current optimization objective function value optimization process function is greater than or equal to 0.001, the current retention times are set to 0.

10. The method for determining parameters of a distributed Raman fiber amplifier according to claim 4, characterized in that: Determining the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, a preset Raman gain coefficient, an optical fiber physical length, an optical fiber attenuation coefficient, an optical fiber effective cross-sectional area, and an optical fiber effective length specifically includes: When the distributed Raman fiber amplifier adopts a forward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression: When the distributed Raman fiber amplifier adopts a backward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression: Among them, G f,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts the forward pumping configuration; G b,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts a backward pumping configuration; t represents the number of pump sources; g R represents the Raman gain coefficient; ω i represents the central angular frequency of the i-th pump source; ω j represents the jth pump source central angular frequency; the pump source central angular frequency is calculated based on the initial optimized pump wavelength and the light propagation speed; A eff represents the effective cross-sectional area of ​​the optical fiber; α represents the attenuation coefficient of the optical fiber; L represents the physical length of the optical fiber; z represents the effective length of the optical fiber; P pj (0) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the forward pump configuration; P pj (L) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the backward pumping configuration.

11. The method for determining parameters of a distributed Raman fiber amplifier according to claim 4, characterized in that: Determining the optimized effective signal-to-noise ratio of each wavelength channel according to the Raman gain, the optical fiber input power of each wavelength channel, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient specifically includes: According to the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient, the cross-phase modulation nonlinear noise coefficient and the fiber-input optical power of each wavelength channel, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel are respectively determined by the following expressions: According to the Raman gain, the optical fiber input power of each of the wavelength channels, the total power loss, the ASE noise, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel, the optimized effective signal-to-noise ratio of each of the wavelength channels is determined by the following expression: Among them, C ∞ represents the correlation coefficient of the self-channel nonlinear noise; η SPM,j represents the self-phase modulation nonlinear noise coefficient; η XPM,ij represents the nonlinear noise coefficient of the cross-phase modulation; P NLI,ci represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel; P ci (0) represents the optical power of the fiber input of the i-th communication channel; P cj (0) represents the optical power of the jth communication channel; P NLI,si represents the nonlinear crosstalk power at the end of the optical fiber of the i-th sensing channel; P si (0) represents the optical power of the i-th sensing channel; SNR link,i represents the signal-to-noise ratio of the i-th wavelength channel; G i A represents the Raman gain of the i-th wavelength channel at each position; i represents the total power loss of the i-th wavelength channel; P i (0) represents the optical power of the i-th wavelength channel; P ASE represents the ASE noise; P NLI,i represents the nonlinear crosstalk power at the end of the optical fiber of the ith communication channel or the sensing channel; SNR eff,i represents the effective signal-to-noise ratio of the i-th wavelength channel; δ and β are both constants.

12. The method for determining parameters of a distributed Raman fiber amplifier according to claim 1, wherein: The method specifically obtains the neural network model through training by the following steps: Based on the initial optimized pump power configuration and the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm, respectively determining the pump power scanning range and the fixed pump wavelength of each of the pump sources; Based on the pump power scanning range and the fixed pump wavelength, using a spectrometer to measure the output optical power of each of the wavelength channels under different fiber input optical powers and pump powers; Constructing a training data set using the input fiber optical power, the pump power and the output optical power; Inputting the training data set into a deep neural network, training the deep neural network with minimizing the loss function as the optimization goal, and obtaining the neural network model; The loss function is specifically: t represents the number of pump sources; represents the actual measured pump power of the i-th pump source; represents the pump power of the i-th pump source output by the deep neural network during each round of training.

13. A parameter determination device for a distributed Raman fiber amplifier, characterized in that: include: An optimization objective function determination module is used to determine the optimization objective function of the link end power configuration of each wavelength channel based on the target application scenario of the optical fiber interawareness integrated system; An optimization objective function solving module, used to solve the optimization objective function using a simulated annealing algorithm to obtain an optimal link end power configuration for each of the wavelength channels; The target pump power determination module is used to obtain the target pump power of each pump source in the distributed Raman fiber amplifier using a pre-trained neural network model according to the fiber input optical power of each wavelength channel and the optimal link end power configuration.

14. The parameter determination device of the distributed Raman fiber amplifier according to claim 13, characterized in that: The optimization objective function determination module is used to determine the optimization objective function of the link end power configuration of each wavelength channel based on the target application scenario of the optical fiber synaesthesia integrated system, and specifically includes: When the target application scenario is a communication function priority scenario, the optimization objective is to maximize the capacity of the communication channel and the effective signal-to-noise ratio of the sensing channel meets the preset minimum effective signal-to-noise ratio requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the communication function priority scenario; When the target application scenario is a sensing function priority scenario, the optimization goal is to maximize the effective signal-to-noise ratio of the sensing channel and the capacity of the communication channel meets the preset minimum capacity requirement, and determine the optimization objective function of the link end power configuration of each wavelength channel in the sensing function priority scenario.

15. The parameter determination device of the distributed Raman fiber amplifier according to claim 14, characterized in that: The optimization objective function is specifically: Where y represents the optimization objective function; n represents the number of communication channels; C i represents the capacity of the i-th communication channel, m represents the number of sensing channels; SNR eff,sj represents the effective signal-to-noise ratio of the jth sensing channel; w1 is used to represent the target application scenario. When w1=1, the target application scenario is the communication function priority scenario; when w1=0, the target application scenario is the sensing function priority scenario.

16. The parameter determination device of the distributed Raman fiber amplifier according to claim 13, characterized in that: The optimization objective function solving module is used to solve the optimization objective function by using a simulated annealing algorithm to obtain the optimal link end power configuration of each wavelength channel, specifically including: Solving the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier; Determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length; Determine the optimized effective signal-to-noise ratio of each of the wavelength channels according to the Raman gain, the optical fiber input power of each of the wavelength channels, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient; The optimal link end power configuration of each of the wavelength channels is determined according to the Raman gain and the optimized effective signal-to-noise ratio.

17. The parameter determination device of the distributed Raman fiber amplifier according to claim 16, characterized in that: The optimization objective function solving module is used to solve the optimization objective function by using a simulated annealing algorithm to obtain an initial optimized pump power configuration and an initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier, specifically including: According to the preset maximum temperature, pump power search space and pump wavelength search space, the temperature, pump power and pump wavelength are randomly initialized; Calculate the updated value of the current pump power and the updated value of the pump wavelength; Determining an updated value of a current optimization objective function value according to the optimization objective function, an updated value of a current pump power, and an updated value of a pump wavelength; According to a comparison result between an updated value of a current optimization objective function value and the current optimization objective function value, the current pump power, the pump wavelength, the optimal pump power, the optimal pump wavelength and the optimal optimization objective function value are updated; Add 1 to the current number of iterations, and update the current temperature based on a preset cooling strategy, and add the current optimal pump power, optimal pump wavelength, and optimal optimization objective function value to the pump power optimization process function, the pump wavelength optimization process function, and the optimization objective function value optimization process function, respectively; re-execute the steps of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of iterations reaches a preset iteration number threshold; wherein the initial value of the number of iterations is 0; Optimizing the process function based on the current optimization objective function value, updating the current number of retention times; resetting the current number of iterations, and re-executing the step of calculating the updated value of the current pump power and the updated value of the pump wavelength until the current number of retention times reaches a preset retention time threshold or the temperature reaches a preset minimum temperature; wherein the initial value of the number of retention times is 0; The initial optimized pump power configuration and the initial optimized pump wavelength configuration of each pump source in the distributed Raman fiber amplifier are obtained according to the optimal pump power added in the current pump power optimization process function and the optimal pump wavelength added in the pump wavelength optimization process function.

18. The parameter determination device of the distributed Raman fiber amplifier according to claim 17, characterized in that: The optimization objective function solving module is used to calculate the updated value of the current pump power and the updated value of the pump wavelength, specifically including: The updated value of the current pump power and the updated value of the pump wavelength are calculated by the following expressions: Among them, x new Indicates the updated value of the current pump power or pump wavelength; x current represents the current pump power or pump wavelength; r represents a random number between -1 and 1; sign() is a sign function; T represents the current temperature; Up and Low are the maximum and minimum values ​​of the pump power search space or the pump wavelength search space, respectively.

19. The parameter determination device of the distributed Raman fiber amplifier according to claim 17, characterized in that: The optimization objective function solving module is used to update the current pump power, pump wavelength, optimal pump power, optimal pump wavelength and optimal optimization objective function value according to the comparison result between the updated value of the current optimization objective function value and the current optimization objective function value, specifically including: When exp(-df / T)>rand(0,1), the current pump power is updated to the updated value of the current pump power, the current pump wavelength is updated to the updated value of the current pump wavelength, and the current optimization objective function value is updated to the updated value of the current optimization objective function value; wherein df is the difference between the updated value of the current optimization objective function value and the current optimization objective function value; T represents the current temperature; rand(0,1) represents a random number between 0 and 1; When exp(-df / T)>rand(0,1) and df>0, the current optimal pump power is updated to the updated value of the current pump power, the current optimal pump wavelength is updated to the updated value of the current pump wavelength, and the current optimal optimization objective function value is updated to the updated value of the current optimization objective function value.

20. The parameter determination device of the distributed Raman fiber amplifier according to claim 17, characterized in that: The optimization objective function solving module is used to update the current temperature based on a preset cooling strategy, specifically including: Update the current temperature using the following expression: T=T max ·exp[-exp(-1)·epoch]。; Where, T represents the current temperature; T max Indicates the maximum temperature; epoch indicates the current iteration number.

21. The parameter determination device of a distributed Raman fiber amplifier according to claim 17, characterized in that: The optimization objective function solving module is used to optimize the process function based on the current optimization objective function value and update the current number of retention times, specifically including: When the difference between the last two items in the optimization process function of the current optimization objective function value is less than 0.001, the current retention times are increased by 1; When the difference between the last two items in the current optimization objective function value optimization process function is greater than or equal to 0.001, the current retention times are set to 0.

22. The parameter determination device of the distributed Raman fiber amplifier according to claim 16, characterized in that: The optimization objective function solving module is used to determine the Raman gain of each wavelength channel at each position according to the initial optimized pump power configuration, the initial optimized pump wavelength configuration, the preset Raman gain coefficient, the optical fiber physical length, the optical fiber attenuation coefficient, the optical fiber effective cross-sectional area and the optical fiber effective length, specifically including: When the distributed Raman fiber amplifier adopts a forward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression: When the distributed Raman fiber amplifier adopts a backward pumping configuration, the Raman gain of each wavelength channel at each position is determined by the following expression: Among them, G f,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts the forward pumping configuration; G b,i(z) represents the Raman gain of each wavelength channel at each position when the distributed Raman fiber amplifier adopts a backward pumping configuration; t represents the number of pump sources; g R represents the Raman gain coefficient; ω i represents the central angular frequency of the i-th pump source; ω j represents the jth pump source central angular frequency; the pump source central angular frequency is calculated based on the initial optimized pump wavelength and the light propagation speed; A eff represents the effective cross-sectional area of ​​the optical fiber; α represents the attenuation coefficient of the optical fiber; L represents the physical length of the optical fiber; z represents the effective length of the optical fiber; P pj (0) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the forward pump configuration; P pj (L) represents the initial optimized pump power of the jth pump source when the distributed Raman fiber amplifier adopts the backward pumping configuration.

23. The parameter determination device of a distributed Raman fiber amplifier according to claim 16, characterized in that: The optimization objective function solving module is used to determine the optimized effective signal-to-noise ratio of each wavelength channel according to the Raman gain, the fiber optical power of each wavelength channel, the preset total power loss, the ASE noise, the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient and the cross-phase modulation nonlinear noise coefficient, specifically including: According to the self-channel nonlinear noise correlation coefficient, the self-phase modulation nonlinear noise coefficient, the cross-phase modulation nonlinear noise coefficient and the fiber-input optical power of each wavelength channel, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel are respectively determined by the following expressions: According to the Raman gain, the optical fiber input power of each of the wavelength channels, the total power loss, the ASE noise, the nonlinear crosstalk power of the optical fiber end of the communication channel and the nonlinear crosstalk power of the optical fiber end of the sensing channel, the optimized effective signal-to-noise ratio of each of the wavelength channels is determined by the following expression: Among them, C ∞ represents the correlation coefficient of the self-channel nonlinear noise; η SPM,j represents the self-phase modulation nonlinear noise coefficient; η XPM,ij represents the nonlinear noise coefficient of the cross-phase modulation; P NLI,ci represents the nonlinear crosstalk power at the end of the optical fiber of the i-th communication channel; P ci (0) represents the optical power of the fiber input of the i-th communication channel; P cj (0) represents the optical power of the jth communication channel; P NLI,si represents the nonlinear crosstalk power at the end of the optical fiber of the i-th sensing channel; P si (0) represents the optical power of the i-th sensing channel; SNR link,i represents the signal-to-noise ratio of the i-th wavelength channel; G i A represents the Raman gain of the i-th wavelength channel at each position; i represents the total power loss of the i-th wavelength channel; P i (0) represents the optical power of the i-th wavelength channel; P ASE represents the ASE noise; P NLI,i represents the nonlinear crosstalk power at the end of the optical fiber of the ith communication channel or the sensing channel; SNR eff,i represents the effective signal-to-noise ratio of the i-th wavelength channel; δ and β are both constants.

24. The parameter determination device of a distributed Raman fiber amplifier according to claim 13, characterized in that: The device also includes a neural network model training module, which is used to: Based on the initial optimized pump power configuration and the initial optimized pump wavelength configuration obtained by solving the optimization objective function using the simulated annealing algorithm, respectively determining the pump power scanning range and the fixed pump wavelength of each of the pump sources; Based on the pump power scanning range and the fixed pump wavelength, using a spectrometer to measure the output optical power of each of the wavelength channels under different fiber input optical powers and pump powers; Constructing a training data set using the input fiber optical power, the pump power and the output optical power; Inputting the training data set into a deep neural network, training the deep neural network with minimizing the loss function as the optimization goal, and obtaining the neural network model; The loss function is specifically: t represents the number of pump sources; represents the actual measured pump power of the i-th pump source; represents the pump power of the i-th pump source output by the deep neural network during each round of training.

25. An electronic device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for determining parameters of a distributed Raman fiber amplifier according to any one of claims 1 to 12 when executing the computer program.

26. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein when the computer program is executed, the device where the computer-readable storage medium is located is controlled to execute the method for determining parameters of a distributed Raman fiber amplifier according to any one of claims 1 to 12.

Images