CN102510307B - Optical fiber disturbance system polarization control method and control system based on annealing algorithm - Google Patents

Abstract

A polarization control method and control system of an optical fiber perturbation system based on an annealing algorithm. In the polarization control, the simulated annealing algorithm is used as the control algorithm, and the polarization controller is used to modulate the polarization state of the interference light wave. Taking the correlation between the two interference signals received by the detector in the distributed optical fiber disturbance location system as the feedback signal, the simulated annealing algorithm is used to search for the corresponding applied voltage value of the polarization controller when the difference between the two signals is the smallest. The polarization control system includes: a continuous distributed optical fiber sensing system based on dual Mach-Zehnder optical fiber interferometers, a squeeze-type polarization controller,

Birefringent phase modulator, single-chip microcomputer system, computer and algorithm software. The optical fiber interference polarization state control method proposed in this method can effectively improve the anti-polarization fading ability of the system by controlling and adjusting the polarization state of the interference light, and largely eliminate the influence of single-mode fiber birefringence on the positioning accuracy of the system.

Description

Polarization control method and control system of optical fiber perturbation system based on annealing algorithm

technical field

The invention belongs to the technical field of sensing and detection.

Background technique

With the development of science and technology and the enhancement of people's security awareness, it has become a necessary and urgent problem to develop a perimeter security system with large detection range, low energy consumption and low cost. Due to its high sensitivity, anti-electromagnetic interference, and no power supply, the distributed optical fiber vibration sensing system has been widely used in military defense, financial protection, energy security, community security and other security fields, and will have a wide range of applications in the future. application prospects.

One of the necessary conditions for light waves to interfere is that the light vectors participating in the interference have the same vibration direction, that is, they have the same polarization direction components. The optical fiber used in the actual perturbation system is an ordinary single-mode optical fiber. Due to the birefringence characteristics of the single-mode optical fiber, the polarization state of the light wave will change after entering the optical fiber, resulting in the degradation of the polarization state of the linearly polarized light. In the field of sensing, this will lead to the reduction of the visibility of the interference fringes, or even the disappearance of the interference fringes, thus greatly reducing the positioning accuracy of the system for disturbances.

Although the use of polarization-maintaining fiber can keep the polarization state of light unchanged, since the laying distance of the fiber in the distributed fiber-optic vibration sensing system is generally tens of kilometers, the use of polarization-maintaining fiber will make the system cost too high. Therefore, in practical applications, we must take another method to combat the polarization degradation of light in the system and compensate for the change of the polarization state of light, so as to eliminate the inaccurate positioning of the system caused by the birefringence characteristics of the single-mode fiber as much as possible.

Contents of the invention

The purpose of the present invention is to solve the problem of inaccurate system positioning caused by the birefringence characteristics of the single-mode optical fiber in the existing method, and to provide a polarization control method and a polarization control system for an optical fiber perturbation system based on a simulated annealing algorithm. The polarization state of the signal light in one arm of the fiber optic vibration sensing system can keep the polarization state of the signal light in the two arms of the system as consistent as possible, thereby improving the positioning accuracy of the system.

Specific technical solutions:

1. Basic principle of distributed optical fiber vibration sensing system

The distributed optical fiber disturbance location system is shown in Figure 1. The system is based on the principle of dual Mach-Zehnder fiber optic interferometers, using two single-mode fibers in the cable to form two test fibers of the Mach-Zehnder fiber optic interferometer to sense the disturbance signal.

The two sensing fibers transmit two groups of light waves in opposite directions at the same time. The disturbance around the optical cable can modulate the phase of the light waves propagating in the fiber, thereby modulating the interference signal. The two beams of light modulated by the phase interfere in the coupler. The interference light is output to the photodetector through the circulator. Since the distance between the location where the disturbance occurs and the photodetectors at both ends of the distributed sensor is different, and the propagation speed of light waves in the optical fiber is constant, the occurrence of the event can be accurately located according to the time difference between the two photodetectors detecting the same event location. The positioning principle is shown in Figure 2. Let the two photodetectors D1 and D2 of the distributed optical fiber disturbance location system detect the same disturbance event as t ₁ and t ₂ respectively, Δt=t ₁ -t ₂ , L is the length of the sensing optical cable, and x is the disturbance Point distance from the position of the second coupler 5, its positioning formula is

$\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; L</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; v</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span>}{\mathrm{<span\; class="notranslate">\; 2</span>}}$

In the formula, v is the propagation velocity of the light wave in the single-mode fiber, in m/s, where v=c/n, c is the speed of light in vacuum (3×10 ⁸ m/s), and n is the refraction of the fiber Rate.

2. Polarization control method of optical fiber perturbation system based on simulated annealing algorithm

The invention adopts the simulated annealing algorithm as the control algorithm to control the extruded optical fiber polarization controller to control the polarization state of the optical signal. In the polarization control, the two signals received by the two detectors in the distributed optical fiber vibration sensing system are used as input , using the simulated annealing algorithm to search for the value of the applied voltage of the polarization controller corresponding to the minimum difference between the two signals. This method proposes a method for controlling the polarization state of interference light in the optical fiber, which can effectively improve the anti-polarization degradation capability of the system, and largely eliminate the influence of single-mode optical fiber birefringence on the disturbance positioning accuracy of the system.

The implementation steps of this method are:

1. In the optical fiber vibration sensing system, the optical signal transmitted in one arm of the sensing fiber is input to the polarization controller, and the interference light of this polarization control and the other interference light are respectively entered into two photodetectors, Use the data acquisition card to collect two signals and input them into the computer;

2. The computer uses the simulated annealing algorithm to judge the difference of the contrast of the two signals to adjust the polarization state, that is, to change the applied voltage value of the extruded polarization controller, to realize continuous control of the signal light passing through the polarization controller, and to use the feedback signal to control the polarization state. The search for the global optimal value of the algorithm is terminated until the correlation degree of the two signals reaches the search termination requirement; by adjusting the polarization state of the interfering light in the system through this method, the polarization error of the data can be effectively suppressed, thereby improving the positioning of the system for disturbances precision.

simulated annealing algorithm

The iterative process of the simulated annealing algorithm is the slow cooling process of the material in the simulated annealing process. The cooling process of annealing is a process in which the internal energy of the object is slowly reduced. From a thermal point of view, the entire internal energy reduction process can be equivalently regarded as a series of equilibrium state transitions. That is to say, the internal energy of the object slowly decreases to zero through a series of specific temperature points from a higher state. There is a heat exchange process between different temperature points, and each temperature point is a quasi-equilibrium state. The annealing algorithm is suitable for the optimization process in the multi-dimensional space, has better global optimal value search performance, and also has a higher convergence speed.

The feedback control of polarization control is actually to optimize the objective function of light intensity in the wave plate phase shift solution space. Based on this, the present invention contemplates using a simulated annealing algorithm to control the polarization state of the retardation-controlled polarization controller. In the solution space composed of multiple wave plate delays, there are many saddle points in the output polarization state light intensity of a fixed direction. The global convergence of the simulated annealing algorithm should be able to effectively solve the false convergence problem of the original gradient method.

In the polarization control, the difference between the two signals received by the two detectors in the distributed optical fiber vibration sensing system (the smaller the difference between the two signals, the better the similarity and the higher the correlation) is used as the feedback signal, that is, the algorithm Therefore, the optimal solution of the objective function is the applied voltage value of the polarization controller corresponding to the minimum difference between the two signals. The difference between the two signals tends to be minimized under the action of the algorithm. When the difference between the two signals meets the iteration termination condition, the search is stopped, and the voltage value corresponding to the best solution is written into the polarization controller, thereby completing the polarization control process.

The flow of the polarization control method is as follows:

First, apply a sine wave with a certain frequency and amplitude to the phase modulator on one arm of the sensor system as a reference signal.

Second, judge whether the difference between the two signals received by the two detectors in the distributed optical fiber disturbance location system is greater than the set threshold. If the difference is greater than the threshold, the computer runs the program and uses the simulated annealing algorithm to control Polarization control is achieved by searching for the optimum voltage value on the polarizer.

Two squeezers of the polarization controller used in polarization control, the problem can be described by the following function:

C=f(V ₁ , V ₂ )

Among them, V _i , i=1, 2 are the voltage values applied on the two extruders, and C is the correlation degree of the two signals received by the two photodetectors in the corresponding positioning system at this time; when C reaches the maximum When the V _i value is the best voltage value, the system is in the best working state at this time, so that the polarization control is realized and the positioning accuracy of the system is improved;

The search process of the simulated annealing algorithm as polarization control is as follows:

(1) Set the initial temperature T ₀ , the end temperature T _e , the attenuation coefficient α, and the optimal value C _m of the two-way signal correlation. The settings of the initial temperature T ₀ , the end temperature T _e , and α need to take into account the global search Performance and calculation time, and α must satisfy the condition: 0<α<1; the state of the polarization controller is described by vector X=(V ₁ , V ₂ ), which represents the optimization problem C=f(V ₁ , V ₂ ) A solution vector of , randomly generate an initial solution vector X ₀ = (V ₁₀ , V ₂₀ ), and assign the correlation C ₀ of the two signals corresponding to the initial solution vector X ₀ to the current optimal solution vector X _b With the current optimal correlation degree C _b , set the current annealing temperature T=T ₀ ;

(2) Judging the size of C _b and C _m , if C _b ≤ C _m , then end the cycle, output the optimal solution vector X _b and the optimal correlation C _b ; if C _b > C _m , then execute step (3 );

(3) Set the number of inner loops n, and set the initial value i of the loop to 1; take the current optimal solution X _b as the center of the circle, and determine according to (V _1k -V _1b ) ² +(V _2k -V _2b ) ² ≤ R ² A circular domain with a radius of R, take an arbitrary solution X _k in the circular domain, and obtain the correlation degree C _k of the two signals corresponding to X _k ;

(4) Judging the size of C _k and C _b , if C _k ≤ _{C b} , update the optimal solution vector X _b and the current optimal correlation C _b so that X _b = X _k and C _b = C _k , and skip Go to step (6); if C _k > C _b , then execute step (5);

(5) Calculate the probability P=exp(-ΔC/T), where ΔC=C _k -C _b , T is the current annealing temperature; randomly generate a number a between 0:1, if P≥a, then accept X _k is the new optimal solution X _b , otherwise continue to iterate with the state X _b ; update the number of inner loops i=i+1; if i<n, return to step 2; if i≥n, update the function T _k according to the temperature ₊₁ = αT _k update the current temperature so that T = αT;

(6) Judge the relationship between the current temperature T and the termination temperature T _e , when T ≥ T _e , return to step 2; when T < T _e , the program terminates, and output the optimal solution vector X _b and the optimal correlation C _b .

3. A control system for realizing the above polarization control method, comprising:

Basic distributed fiber optic sensor based on dual Mach-Zehnder fiber optic interferometers: used to generate interference signals and perform disturbance positioning. The polarization control system is realized by adding polarization control devices on the basis of the sensor;

Squeeze-type polarization controller: it has four squeezers arranged staggered at 45° in the extrusion direction. The first two fiber squeezers are used in the control process. Applying different combinations of voltages to the two squeezers can control The polarization state of the input light wave is modulated differently to output light waves of different polarization states; a polarization controller is added to one of the sensing fibers of the above-mentioned basic distributed optical fiber sensor, and the polarization control is realized by modulating the polarization state of the optical signal of the path;

LiNbO ₃ birefringent phase modulator: used to generate a reference signal, added to the other side of the sensing fiber corresponding to the end of the basic distributed optical fiber sensor and the polarization controller, used to generate sinusoidal phase modulation, and the phase modulation signal passes through the dual Mach- Zehnder fiber optic interferometer interference produces intensity sinusoidal modulation, which is received by two photodetectors separately. If there is no polarization degradation, the two interference signals received by the two photodetectors in the basic distributed fiber optic sensor have a fixed time delay and equal amplitude. ;

Data acquisition card: collect the voltage signals of two photodetectors and send them to the computer for processing;

Computer: through the software programming in the computer, the processing of the acquisition signal sent by the data acquisition card is realized, so as to realize the iterative search of the optimal modulation voltage, and the optimal modulation voltage searched is fed back to the polarization controller and the polarization controller through the single-chip microcomputer system phase modulator;

Single-chip microcomputer system: through communication with the computer, the output digital signal directly controls the polarization controller; the output sine wave signal modulates the phase modulator.

The polarization control system is shown in Figure 5.

Advantages and beneficial effects of the present invention:

At present, most distributed optical fiber disturbance location systems have poor positioning accuracy due to lack of polarization control. The invention proposes a polarization control method used in a distributed optical fiber disturbance location system. Using this method can effectively improve the anti-polarization degradation capability of the system, and largely eliminate the influence of single-mode fiber birefringence on the disturbance positioning accuracy of the system. The simulated annealing algorithm used in this method is a method for finding the global optimal solution in nonlinear programming problems. It has good global optimal value search performance and high convergence speed, and can effectively solve The original gradient-based false convergence problem.

Description of drawings

Figure 1 is a distributed optical fiber disturbance location system;

Fig. 2 is a positioning principle diagram;

Fig. 3 is two signals without polarization control collected by the distributed optical fiber disturbance location system;

Figure 4 shows two signals collected by the system after polarization control;

Fig. 5 is a polarization control system;

Fig. 6 is a schematic diagram of polarization control algorithm search;

In the figure, 1 and 9 are lasers, 2 and 10 are first couplers, 5 and 13 are second couplers, 6 and 14 are third couplers, 3 and 11 are first optical circulators, 4 and 12 are The second optical circulator, 7a is the sensing fiber F1 in the optical cable, 7b is the sensing fiber F2 in the optical cable, 8a and 18a are the first photodetector D1, 8b and 18b are the second photodetector D2, 15 is the polarization control device, 16 is a LiNbO ₃ birefringence phase modulator, 17 is a sensing optical cable, 19 is a single-chip microcomputer system, 20 is a computer, and 21 is an acquisition card.

Detailed ways:

Embodiment 1: A polarization control system used in a distributed optical fiber disturbance location system

As shown in Figure 5, the system includes:

Laser 9, first coupler 10, second coupler 13 and third coupler 14, first optical circulator 11 and second optical circulator 12, sensing optical cable 17, first photodetector 18a and second photoelectric Detector 18b.

Squeeze-type polarization controller 15: It has four squeezers arranged staggered at 45° in extrusion directions. The first two fiber squeezers are used in the control process, and different combinations of voltages can be applied to the two squeezers. The polarization states of the input light waves are modulated differently to output light waves of different polarization states. Add a polarization controller to one of the sensing fibers of the basic distributed optical fiber sensing system, and realize polarization control by modulating the polarization state of the optical signal of the path;

LiNbO ₃ birefringent phase modulator 16: used to generate a reference signal, added to the other path of the sensing fiber corresponding to the end of the polarization controller in the basic distributed optical fiber sensing system, used to generate sinusoidal phase modulation, and the phase modulation signal passes through The interference of double Mach-Zehnder fiber optic interferometers produces intensity sinusoidal modulation, which is received by two photodetectors respectively. If there is no polarization degradation, the two interference signals received by the two photodetectors in the basic distributed optical fiber sensing system have a fixed time extended and equal in magnitude;

Data Acquisition Card (DAQ Card) 21: collect the voltage signals of the two photodetectors 18a and 18b, and send them to the computer for processing.

Computer (PC) 20: realize the processing of the acquisition signal that data acquisition card is sent into by the software programming in computer, to realize the iterative search of optimum modulation voltage, and the optimum modulation voltage searched is fed back to by single-chip microcomputer system Polarization controller and phase modulator.

Single-chip microcomputer system (SCM) 19: communicate with the computer, output digital signal to directly control the polarization controller; output sine wave signal to modulate the phase modulator.

Embodiment 2: Polarization control method

One of the important conditions for two beams of light to interfere is that the light participating in the interference is linearly polarized and has the same polarization direction. The positioning algorithm of the distributed optical fiber disturbance location system is based on the assumption that linearly polarized light interferes with the same polarization direction.

Distributed optical fiber disturbance positioning system is mainly used for perimeter protection, earthquake monitoring, etc. The optical fiber used is tens of kilometers or even hundreds of kilometers long. If the polarization maintaining optical fiber and its corresponding supporting components are used, it will be very expensive. Considering the cost, the system The sensing fibers used are all commonly used single-mode fibers.

Ordinary single-mode optical fibers will cause birefringence due to random factors such as geometric bending, ambient temperature changes, and other non-random errors, which will cause the polarization state of linearly polarized light to change when it is transmitted in ordinary single-mode optical fibers, resulting in linearly polarized light. polarization degradation. Polarization state degradation will not only affect the quality of the interference output signal, but also seriously affect the positioning accuracy of the entire system.

Figure 3 shows the two signals collected by the distributed optical fiber disturbance location system without polarization control. The two signals shown in the figure are very different, and the connection between the two signals is completely invisible. Disturbance is applied to the point at position 128.6m, the positioning results are randomly distributed from 174.9m to 360m, the maximum deviation from the actual position reaches 231.4m, and the average deviation reaches 175.9m, which completely deviates from the actual position.

The state of the two signals after the polarization state is adjusted is shown in FIG. 4 , and the two signals are basically the same at this time. Similarly, the disturbance at the position 128.6m is located. The positioning result is from 123.4m to 133.7m. The maximum deviation from the actual position is 5.2m, and the average deviation is 1.6m. The positioning accuracy has increased significantly.

Therefore, in the distributed optical fiber disturbance positioning system for practical application, it is necessary to control the polarization of the interference light and compensate the change of the polarization state of the interference light, so as to improve the anti-polarization fading ability of the whole system and improve the positioning accuracy of the system.

As shown in Figure 5, the optical signal of one arm of the sensing fiber is input into the polarization controller 15, and after polarization control, the two interference signals respectively enter the two photodetectors, and the data acquisition card collects the two signals and sends the signals to into the computer.

The annealing algorithm of the computer's internal software adjusts the position vector of each particle in the chaotic particle swarm optimization algorithm according to the correlation of the two-way interference signals fed back, that is, changes the voltage value applied to the extruder of the polarization controller corresponding to each position vector. Continuously control the polarization state of the incident light wave of the polarization controller and use the feedback signal to search for the optimal value until the correlation of the two signals corresponding to the feedback signal meets the search termination condition.

According to the system shown in Figure 5, combined with the search process shown in Figure 6, the calculation process of the polarization control method based on the annealing algorithm is firstly described, and then the influence of the control method on the positioning accuracy of the system is illustrated through experiments, and finally the annealing algorithm is discussed The value rules of each parameter.

1. In order to explain the calculation process of the annealing algorithm, an example of polarization control is used to illustrate:

(1) Initialize settings. Initial temperature T ₀ =1, end temperature T _e =0.05, temperature attenuation coefficient α=0.5, optimal value for setting the correlation with two signals C _m =0.01, radius R=100, number of inner cycles n= 20. Randomly generate an initial solution vector X ₀ = (1251, 2870), and assign the correlation degree C ₀ = 0.0495831 of the two signals corresponding to the initial solution vector X ₀ to the current optimal solution vector X _b and the current optimal correlation Degree C _b ;

(2) Since C _b =0.0495831>C _m =0.01 at this time, the termination condition is not satisfied, continue to execute step (3);

(3) After 20 inner loops, X _b = (1390, 2560), the corresponding optimal correlation degree C _b = 0.0339063 is greater than C _m , update the current temperature T = 0.5;

(4) Repeat the inner loop after updating the temperature until after 5 inner loops, X _b = (1282, 1963), the corresponding optimal correlation degree C _b = 0.021536 is still greater than C _m , but at this time the temperature T = 0.03125 <T _e =0.05, the termination condition is satisfied, the program terminates, and the optimal solution vector X _b =(1282, 1963) and the optimal correlation degree C _b =0.021536 are output at this time;

2. The influence of this control method on the positioning accuracy of the system:

Several parameters in the annealing algorithm are set as follows: initial temperature T ₀ =1, end temperature T _e =0.05, temperature attenuation coefficient α=0.5, and optimal value C _m for setting the correlation with the two signals = 0.01, radius R=100, number of inner loops n=20. The polarization state is thereby adjusted.

Table 1 Experimental data on the influence of polarization control based on annealing algorithm on system positioning accuracy (unit: m)

It is known that the position of the disturbance applied in the experiment is 128.6m. From the experimental results, it can be known that when the polarization state is not adjusted, the average value of positioning is 304.5m. It can be seen that the positioning accuracy when the polarization state is not adjusted is above 176m; when the polarization state is adjusted After the state, the average value of the positioning results is 137.8m, and the positioning accuracy at this time is within 10m. It can be found that the positioning accuracy has increased by nearly 130m. It can be seen that the addition of polarization control can make the positioning of the system more accurate for disturbances, thereby optimizing the system.

3. The value rules of each parameter of the annealing algorithm:

In the following, several sets of experiments are used to illustrate the value rules of parameters such as initial temperature T ₀ , temperature decay constant α, radius R, and number of internal cycles n in the simulated annealing algorithm.

(1) Initial temperature T ₀

The initial temperature T ₀ , the termination temperature T _e and the temperature decay constant α are mutually restrictive, so they should be considered comprehensively. In this experiment, T ₀ was changed, and other parameters were selected as follows: T _e =0.01, α=0.5, C _m =0.02, R=100, n=30. The final convergence effect (optimum correlation degree C _b ) and convergence time in the experiment are shown in Table 2 below.

Table 2 Experimental data of convergence effect and convergence time when changing the initial temperature T ₀

C _b 1 2 3 4 5 average t T ₀ =0.2 0.0455441 0.0309543 0.0386342 0.0495051 0.0340284 0.03973322 20s T ₀ =0.5 0.0565112 0.0664027 0.0312667 0.0667577 0.024546 0.04909686 24.8s T ₀ =1 0.024829 0.045174 0.0351302 0.0296336 0.0231083 0.03157502 27s

It can be seen from the experimental data that the greater the initial temperature T ₀ , the better the convergence effect, but the convergence time increases, and even the program cannot run later. Comparing the above experimental results comprehensively, considering the convergence effect and the time used, the experimental effect is the best when the initial temperature T ₀ =1.

(2) Temperature decay constant α

The temperature decay constant α reflects the temperature decay rate. In this experiment, α was changed, and other parameters were selected as follows: T ₀ =1, T _e =0.01, C _m =0.02, R=100, n=30. The convergence effect and convergence time obtained in the experiment are shown in Table 3 below.

Table 3 Experimental data of convergence effect and convergence time when changing temperature decay constant α

C _b 1 2 3 4 5 average t α＝0.2 0.0283965 0.0501971 0.0452508 0.0449461 0.0327374 0.04030558 12.5s α＝0.5 0.024829 0.045174 0.0351302 0.0296336 0.0231083 0.03157502 27s α=1 0.0324687 0.0420882 0.0257653 0.0302355 0.0285679 0.03182512 1min20s

It can be seen from the experimental data that within a certain range, the larger the temperature decay constant α is, the slower the cooling rate is, but the quality of the corresponding solution is higher. Comprehensively comparing the above experimental results, the quality of the solution does not increase significantly when α increases, but the time increases significantly, so it is more appropriate to take α=0.5 in the experiment.

(3) Access radius R

The access radius R in the state update function determines the search range of state X. Therefore, when R is too small, the program cannot search the entire solution space, and the local optimal value will be returned instead of the global optimal value; and when R is too large, the program will fail to run. In this experiment, R was changed, and the choices of several other parameters were: T ₀ =1, T _e =0.01, C _m =0.02, α=0.5, n=30. The convergence effect and convergence time obtained in the experiment are shown in Table 4 below.

Table 4. Experimental data of convergence effect and convergence time when changing the access radius R

C _b 1 2 3 4 5 average t R=10 0.117228 0.0844735 0.117081 0.054626 0.0660967 0.08790104 28.4s R = 50 0.0873674 0.0358196 0.11275 0.0420517 0.0436244 0.06432262 27.8s R=100 0.024829 0.045174 0.0351302 0.0296336 0.0231083 0.03157502 27s

It can be seen from the experimental data that the larger R is, the better the solution is, and the less time it takes, but when R is greater than 100, sometimes it will make the program difficult to run, so R=100 is used in this experiment.

(4) Number of inner loops n

The more n the inner loop performs, the larger the solution space the program searches, and the longer it takes. In this experiment, n was changed, and the selection of several other parameters were: T ₀ =1, T _e =0.01, C _m =0.02, α=0.5, R=100. The convergence effect and convergence time obtained in the experiment are shown in Table 5 below.

Table 5 Experimental data of convergence effect and convergence time when changing the number of inner loops n

C _b 1 2 3 4 5 average t n=20 0.0699163 0.0518094 0.0398634 0.0287355 0.0440567 0.04687626 19.8s n=30 0.024829 0.045174 0.0351302 0.0296336 0.0231083 0.03157502 27s n=40 0.0302194 0.0308112 0.0382326 0.0327123 0.0274071 0.03187652 36.5s

It can be seen from the experimental results that when n is too small, the experimental effect is not good, and when n is too large, the program cannot run. Considering the factors such as convergence effect, time and program quality comprehensively, n=30 is desirable.

Therefore, the optimal values of these parameters obtained through experiments are: T ₀ =1, T _e =0.01, C _m =0.02, α=0.5, R=100, n=30.

Claims (3)

1. be applied to a polarization control method for distributed optical fiber disturbance positioning system, it is characterized in that the implementation of the method is:

In distributed optical fiber disturbance positioning system, the light signal of sensor fibre one arm is input in extrusion pressing type Polarization Controller, after Polarization Control, two-way interference signal enters respectively two photodetectors, and data collecting card gathers two paths of signals and signal is sent into computer;

Computer is adjusted the position vector of each particle in simulated annealing according to the degree of correlation of fed back two-way interference signal, change each position vector corresponding be applied to the magnitude of voltage on the first two optical fiber squeezer on extrusion pressing type Polarization Controller, the polarization state of extrusion pressing type Polarization Controller incident light wave is carried out continuous control and utilized feedback signal to carry out optimal value search, until the degree of correlation of two paths of signals corresponding to feedback signal stops while meeting search end condition.

2. method according to claim 1, is characterized in that the detailed process of described simulated annealing is: two squeezers of the extrusion pressing type Polarization Controller using in Polarization Control, this problem can be described in order to minor function:

C=f(V ₁,V ₂)

V wherein _i, i=1,2 are respectively the magnitude of voltage applying on two squeezers, and C is the degree of correlation of the two paths of signals that in now corresponding navigation system, two photodetectors receive; V when C reaches maximum _ivalue is for optimum voltage value, and now system is in optimum Working, thereby realized Polarization Control, improved the positioning precision of system;

Search procedure as the simulated annealing of Polarization Control is as follows:

(1) initial temperature T is set ₀, final temperature T _e, the setting optimal value C of attenuation coefficient α and the two paths of signals degree of correlation _m, initial temperature T wherein ₀, final temperature T _eand arranging of α need to take into account global search performance and computing time, and α need satisfy condition: 0< α <1; Vectorial X=(V for the state of extrusion pressing type Polarization Controller ₁, V ₂) describe, it represents optimization problem C=f (V ₁, V ₂) a solution vector, generate at random initial solution vector X ₀=(V ₁₀, V ₂₀), and will with initial solution vector X ₀the degree of correlation C of corresponding two paths of signals ₀be assigned to respectively current optimal solution vector X _bwith current optimum degree of correlation C _b, current annealing temperature T=T is set ₀;

(2) judgement C _band C _msize, if C _b≤ C _m, end loop, exports optimal solution vector X _bwith optimum degree of correlation C _b; If C _b>C _m, perform step (3);

(3) interior cycle-index n is set, juxtaposition circulation initial value i is 1; With current optimal solution X _bas the center of circle, according to (V _1k-V _1b) ²+ (V _2k-V _2b) ²≤ R ²determine that Radius is in the ，Yuan territory, round territory of R, to get one to separate arbitrarily X _k, and try to achieve and X _kthe degree of correlation C of corresponding two paths of signals _k;

(4) judgement C _kwith C _bsize, if C _k≤ C _b, upgrade optimal solution vector X _bwith current optimum degree of correlation C _b, make X _b=X _kand C _b=C _k, and jump to step (6); If C _k>C _b, perform step (5);

(5) calculating probability P=exp (△ C/T), wherein △ C=C _k-C _b, T is current annealing temperature; The random several a that produce 0～1, if P>=a accepts X _kfor new optimal solution X _b, otherwise continue with state X _biteration; Cycle-index i=i+1 in upgrading; If i < is n, get back to step (2); If i>=n, according to temperature renewal function T _k+1=α T _kupgrade Current Temperatures and make T=α T;

(6) judgement Current Temperatures T and final temperature T _erelation, as T>=T _etime, get back to step (2); Work as T<T _etime, program stops, output optimal solution vector X _bwith optimum degree of correlation C _b.

3. realize a polarization control system for method claimed in claim 1, it is characterized in that this polarization control system comprises:

Basic distributed fiberoptic sensor based on two Mach-Zehnder fibre optic interferometers: for generation of interference signal, carry out disturbances location, polarization control system is to add what extrusion pressing type Polarization Control device was realized on the basis of this transducer;

Extrusion pressing type Polarization Controller: have four direction of extrusion staggered squeezers at 45 °, in control procedure, use the first two optical fiber squeezer wherein, the voltage that two squeezers is applied to various combination can carry out different modulation to the polarization state of input light wave, thus the light wave of output different polarization states; Extrusion pressing type Polarization Controller is added in to a wherein road of above-mentioned basic distributed fiberoptic sensor sensor fibre, by modulating the polarization state of this road light signal, realizes Polarization Control;

LiNbO ₃birefringent phase modulator: for generation of reference signal, be added in another road of the sensor fibre corresponding with extrusion pressing type Polarization Controller place end of basic distributed fiberoptic sensor, for generation of sinusoidal phase modulation, phase modulated signal is interfered and is produced intensity Sine Modulated through two Mach-Zehnder fibre optic interferometers, by two photodetectors, received respectively, if do not exist polarization to degenerate, the two-way interference signal that in basic distributed fiberoptic sensor, two photodetectors receive has fixed delay and amplitude equates;

Data collecting card: the voltage signal to two photodetectors gathers, and send into computer and process;

Computer: realize the processing of the collection signal that data collecting card is sent into by the software programming in computer, to realize the iterative search of best modulation voltage, and the best modulation voltage searching is fed back to extrusion pressing type Polarization Controller and phase-modulator by Single Chip Microcomputer (SCM) system;

Single Chip Microcomputer (SCM) system: by communicating with computer, output digit signals is directly controlled extrusion pressing type Polarization Controller; Sine wave output signal, modulates phase-modulator.

Images