CN105719003A - Quantum genetic algorithm-based converter transformer partial-discharge ultrasonic location method - Google Patents

Abstract

The invention provides a quantum genetic algorithm-based converter transformer partial-discharge ultrasonic location method. Many ultrasonic sensors at different positions of a transformer are adopted to receive ultrasonic signals that are sent by a partial discharge source, and a distance solving model is established by a Cartesian coordinate system. Relevant parameters of the quantum genetic algorithm are initialized. Chromosome is coded, and population Q(t) is initialized. Every individual of the initialized population is measured and a state P(t) is obtained. The fitness to each state is calculated. The optimal individual and its fitness value are recorded. The result is directly output if the termination condition is satisfied. If the termination condition is not satisfied, then t=t+1 is set and every individual of the initialized population is measured to obtain the state P(t). The fitness to each state is calculated, the population individuals are updated by means of quantum rotation gate operation and quantum non-gate to obtain a progeny population Q(t+1), and the optimal individual and its fitness are recorded, until the terminal condition is satisfied. The method of the invention has the characteristics that the iteration frequency is low and the location precision is high under the condition of a small population size, premature convergence is avoided, and rapid convergence is achieved to get the globally optimal solution.

Description

A kind of converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm

Technical field

The invention belongs to converter power transformer on-line monitoring technique field, particularly relate to a kind of converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm.

Background technology

Converter power transformer is the important electrical in DC transmission system, is in hinge status in DC transmission system, and therefore the health status of converter power transformer is paid much attention in power system, especially the health status of its dielectric.Along with the raising of the development of power system and electric pressure, shelf depreciation has become as the major reason of converter power transformer insulation degradation, thus the detection of shelf depreciation and location also just become the important means of its insulation status monitoring.

The ultrasonic locating of converter power transformer shelf depreciation, in order to get rid of electrical Interference during on-the-spot transformer fault location, the several hyperboloid positioning modes of many employings.The method adopts acoustical signal triggering system, namely in the one group of array of ultrasonic sensors being distributed in fuel tank side, by receiving the sensor of ultrasonic signal at first as trigger source, triggers all the other sound passages.Owing to the interference in Sound Surveillance System is relatively fewer, therefore the situation of disturbing pulse false triggering in electro-acoustic location can be avoided to occur.Select trigger source sensor as reference sensor during location, as benchmark, measure the relative time delay corresponding to it when same acoustic emission signal propagates all the other each sensors, these relative time delays are substituted into the one group of hyperboloid solving equations meeting this array geometry relation, the geometric position of discharge source anchor point can be tried to achieve.

Based on the basic theories of above-mentioned location, propose both at home and abroad at present and multiple realize method, such as genetic algorithm, unit module searching algorithm, pattern recognition and innovatory algorithm thereof etc..All kinds of methods all have its feature, and speed and precision to location Calculation have raising in various degree.Based on the localization method of genetic algorithm, compared with other algorithms, having program structure simple, to the initial point requirement without feasibility, the dimension of problem is unrestricted advantage also.But traditional genetic algorithm exists local search ability difference in theory and practice and the defect of Premature Convergence easily occurs, and for the optimization problem of some complexity always, it is easy to fall into locally optimal solution, and the solution of global optimum can not be reached.

Summary of the invention

The present invention mainly solves the technical problem existing for prior art: provide a kind of converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm；Compared with traditional genetic algorithm, quantum genetic algorithm under less population scale, can converge to globally optimal solution rapidly, and the introducing of quantum door makes algorithm possess development ability and exploring ability, it is ensured that algorithmic statement.

The above-mentioned technical problem of the present invention is addressed mainly by following technical proposals:

Step one: adopt the sonac being placed in the multiple diverse location of transformator to receive the ultrasonic signal that Partial Discharge Sources sends, set up distance solving model with cartesian coordinate system；

Step 2: determine quantum genetic algorithm initial parameter, arranges maximum population G and population scale N from generation to generation, chromosome is encoded, makes t=0, initialize population Q (t)；

Step 3: to each individual enforcement one-shot measurement of initial population Q (t), obtain a state P (t)；

Step 4: this group solution of state P (t) is carried out Fitness analysis, records the optimal adaptation degree individuality desired value as next step evolution；

Step 5: each state computation fitness to initial population；

Step 6: if end condition is unsatisfactory for, then make t=t+1, then according to certain adjustable strategies, utilizes Quantum rotating gate operation that population at individual is updated, obtains progeny population Q (t+1), repeat step 3, step 4 and step 5；Above-mentioned described end condition, according to required precision needed for target solution, is set；

Step 7: record optimized individual and fitness, until end condition meets, stops algorithm output result.

Preferably, in above-mentioned steps one, the quantity of sonac is that n (n >=6) is individual, is arranged on converter power transformer fuel tank outer wall.

Under fuel tank, cartesian coordinate system is set up for zero in summit, sets up nonlinear mathematical model:

$\left\{\begin{array}{c}\sqrt{{(x-x_2)}^2+{(y-y_2)}^2+{(z-z_2)}^2}-\sqrt{{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2}=v_s{\mathrm{\τ}}_{12}\\ \sqrt{{(x-x_3)}^2+{(y-y_3)}^2+{(z-z_3)}^2}-\sqrt{{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2}=v_s{\mathrm{\τ}}_{13}\\ ......\\ \sqrt{{(x-x_n)}^2+{(y-y_n)}^2+{(z-z_n)}^2}-\sqrt{{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2}=v_s{\mathrm{\τ}}_{1n}\end{array}\right.$

Set a trap to put and be a little positioned at that (x, y, in z), sensor location coordinates is followed successively by (x₁,y₁,z₁)、(x₂,y₂,z₂)……(x_n,y_n,z_n), equivalent velocity of sound v_sIf, with sensor (x₁,y₁,z₁) for trigger source (real system is the sensor receiving ultrasonic signal at first as trigger source, be herein for the ease of illustrate), τ_1i, i=2,3 ..., n is that each sensor is with (x, y, the several signal time delay between z), namely sound wave passes to the time difference between different sensors.

Sound wave can reflect in inside transformer, reflects, diffraction etc., and the spread speed difference in other solid insulating materials of coil neutralization is very big, therefore velocity of wave is not a constant, if v_s∈ [1,1.6], (mm/ μ s), wherein mm/ μ s represents the every microsecond of millimeter.Depending on v_sThe unknown, in institute's above formula, parameter to be asked is x, y, z, v_s.Above-mentioned equation group is represented by:

f_i(x,y,z,v_s)

=[(x-x_i)²+(y-y_i)²+(z-z_i)²]^1/2-[(x-x₁)²+(y-y₁)²+(z-z₁)²]^1/2-v_sτ_1i=0

In order to position is put in location office more accurately, adding constraints, above formula becomes:

$\begin{array}{c}min\left\{D_f(x,y,z,v_s)\right\}=\\ {\mathrm{\Σ}}_{i=2}^n{\{{\[{(x-x_i)}^2+{(y-y_i)}^2+{(z-z_i)}^2\]}^{1/2}+{\[{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2\]}^{1/2}-v_s{\mathrm{\τ}}_{1i}\}}^2\end{array}$

Constraints is:

$\left.\begin{array}{c}0\≤x\≤x_{\max }\\ 0\≤y\≤y_{\max }\\ 0\≤z\≤z_{\max }\\ 1\≤v_s\≤1.6,(mm/\mathrm{\μ}s)\end{array}\right\}$

The length coordinate of a position, v are put in x, y, z respectively office_sFor converter power transformer internal ultrasonic ripple equivalence velocity of wave, x_max,y_max,z_maxFor the actual length of converter power transformer tank wall；

Above formula is transcendental equations, it is impossible to directly obtain Exact Solutions, optimal solution when obtaining physical constraint only with computerized algorithm；

Preferably, chromosome carrying out in described step 2 quantum bit coding, each quantum individuality is with following quantum form coding:

$q_j^t=\[\left.\begin{array}{c}{\mathrm{\α}}_{11}^t\\ {\mathrm{\β}}_{12}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{12}^t\\ {\mathrm{\β}}_{12}^t\end{array}\right|\left.\begin{array}{c}\mathrm{...}\\ \mathrm{...}\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{1k}^t\\ {\mathrm{\β}}_{1k}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{21}^t\\ {\mathrm{\β}}_{21}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{22}^t\\ {\mathrm{\β}}_{22}^t\end{array}\right|\left.\begin{array}{c}\mathrm{...}\\ \mathrm{...}\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{2k}^t\\ {\mathrm{\β}}_{2k}^t\end{array}\right|\left.\begin{array}{c}\mathrm{...}\\ \mathrm{...}\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{n1}^t\\ {\mathrm{\β}}_{n1}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{n2}^t\\ {\mathrm{\β}}_{n2}^t\end{array}\right|\left.\begin{array}{c}\mathrm{...}\\ \mathrm{...}\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{nk}^t\\ {\mathrm{\β}}_{nk}^t\end{array}\right|\]$

WhereinBeing that t is individual for jth, n is the quantum gene number that each quantum is individual, and k is the quantum bit number used by the component of each independent variable, α_ij,β_ijIt is that two plural numbers are called probability amplitude pair, and meets: | α_ij|²+|β_ij|²=1, | α_ij|²For quantum be in spin downward probability of state, | β_ij|²It is in for quantum and spins up probability of state.

Preferably, in described step 3, to each individual enforcement one-shot measurement of initial population Q (t), obtain a state P (t)；During measurement, it is according to quantum bit probability amplitude | α_ij|²(or | β_ij|²) select 0 on corresponding gene position or 1.

Method particularly includes: randomly generate one [0,1], if it is be more than or equal to probability amplitude | α_ij|²(or | β_ij|²) value, then measurement result takes 1；Otherwise taking 0, vice versa.Thus the individuality of quantum coding is converted to binary-coded individuality, obtains N number of binary-coded individuality.

Preferably, in described step 4, described in solve fitness and refer to and utilize the fitness of binary coding solved function obtained, in Numerical Optimization, process is: first binary code is converted to decimal number, then substitute in function to be optimized, obtain its functional value and be fitness；The function herein optimized and fitness function, fitness function is used for judging the superiority-inferiority of individuality or solution, chooses different fitness functions according to different target function when specifically solving, and fitness function has a variety of, can directly invoke in MATLAB workbox, it is possible to oneself structure as required.

Preferably, in described step 6, the renewal of quantum bit is by more having newly arrived to quantum door, and its process is:

$\left[\begin{array}{c}{\mathrm{\α}}_i^{\′}\\ {\mathrm{\β}}_i^{\′}\end{array}\right]=\left[\begin{array}{cc}\cos \left({\mathrm{\θ}}_i\right)& -\sin \left({\mathrm{\θ}}_i\right)\\ \sin \left({\mathrm{\θ}}_i\right)& \cos \left({\mathrm{\θ}}_i\right)\end{array}\right]\left[\begin{array}{c}{\mathrm{\α}}_i\\ {\mathrm{\β}}_i\end{array}\right]$

Wherein [α_i,β_i]^TIt is the quantum bit of i-th chromosome, θ_iFor the anglec of rotation.

Described adjustable strategies is as shown in the table:

Wherein x_iFor the i-th bit of current chromosome, BEST_iFor the current chromosomal i-th bit of optimum, f (x) is fitness function, Δ θ_iFor anglec of rotation size, the convergence rate of control algolithm, s (α_i,β_i) it is the direction of the anglec of rotation, it is ensured that convergence of algorithm, and meet θ_i=s (α_i,β_i)Δθ_i。

The present invention is compared with prior art, have the advantages that, converter power transformer type local-discharge ultrasonic location algorithm based on quantum genetic algorithm has positioning precision height, avoid Premature Convergence and under less population scale, the feature of globally optimal solution can be converged to rapidly, the introducing of quantum door makes algorithm possess development ability and exploring ability, it is ensured that algorithmic statement.

Accompanying drawing explanation

Fig. 1 is that source position and fuel tank outer wall sensing station schematic diagram are put in office；

Xoyz is by being built cartesian coordinate system, and (x, y z) put source position, S for office to P₁(x₂,y₂,z₂),S₂(x₂,y₂,z₂),S₃(x₃,y₃,z₃),S₄(x₄,y₄,z₄),S₅(x₅,y₅,z₅),S₆(x₆,y₆,z₆) respectively fuel tank outer wall sonac position；

Fig. 2 is converter power transformer type local-discharge ultrasonic location quantum genetic algorithm flow chart.

Detailed description of the invention

Below in conjunction with the drawings and specific embodiments, present disclosure is described in further details.

Embodiment:

Phantom is set up for certain converter power transformer shelf depreciation field experimentation Monitoring Data, utilize quantum genetic algorithm and carry out fault location, and compare with traditional genetic algorithm, result proves quantum genetic algorithm, and to have positioning precision higher, it is to avoid Premature Convergence and converge to the feature of globally optimal solution rapidly.

The present invention comprises the steps: based on the converter power transformer type local-discharge ultrasonic localization method of quantum genetic algorithm

Step one: adopt the sonac being placed in the multiple diverse location of transformator to receive the ultrasonic signal that Partial Discharge Sources sends, set up distance solving model with cartesian coordinate system；Certain converter power transformer fuel tank actual size is 5m × 4m × 3m, and source coordinate is put for (4.500,3.700,2.600) in actual office, sonic sensor coordinate respectively S₁(2.500,2.500,0), S₂(0,2.0000,1.500), S₃(2.500,2.000,3.000), S₄(5.000,2.000,1.500), S₅(2.500,4.000,1.500), S₆(4.000,0,2.000)；With S₁Sensor-triggered source, S₁(2.500,2.500,0) are reference coordinate, t₁=2612 μ s are the reference time, then several time delays respectively τ of other accepted signals of each sensor_1i(μ s)=[3525,1958,1523,1644,2711]-t₁, i=2,3,4,5,6, data above is substituted into following formula:

$\begin{array}{c}min\left\{D_f(x,y,z,v_s)\right\}=\\ {\mathrm{\Σ}}_{i=2}^n{\{{\[{(x-x_i)}^2+{(y-y_i)}^2+{(z-z_i)}^2\]}^{1/2}+{\[{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2\]}^{1/2}-v_s{\mathrm{\τ}}_{1i}\}}^2\end{array}$

Constraints:

$\left.\begin{array}{c}0\≤x\≤5.000\\ 0\≤y\≤4.000\\ 0\≤z\≤3.000\\ 1\≤v_s\≤1.6,(mm/\mathrm{\μ}s)\end{array}\right\}$

Step 2: determine quantum genetic algorithm initial parameter, maximum population G and population scale N from generation to generation is set, chromosome is encoded, makes t=0, initialize population Q (t), make maximum population G=200 from generation to generation, population scale N=40, in described initialization of population, population scale is N=40, namely having 40 quantum coding individualities, each quantum individuality is set toNamely

$\left[\begin{array}{c}\frac{1}{\sqrt{2}}\frac{1}{\sqrt{2}}...\frac{1}{\sqrt{2}}\\ \frac{1}{\sqrt{2}}\frac{1}{\sqrt{2}}...\frac{1}{\sqrt{2}}\end{array}\right]$

Step 3: to each individual enforcement one-shot measurement of initial population Q (t), obtain a state P (t)；During measurement, it is according to quantum bit probability amplitude | α_ij|²(or | β_ij|²) select 0 on corresponding gene position or 1. method particularly includes: randomly generate one [0,1], if it is be more than or equal to probability amplitude | α_ij|²(or | β_ij|²) value, then measurement result takes 1；Otherwise taking 0, vice versa.Thus the individuality of quantum coding is converted to binary-coded individuality, obtains 40 binary-coded individualities.

Step 4: this group solution is carried out Fitness analysis, records the optimal adaptation degree individuality desired value as next step evolution.The described fitness that solves refers to the fitness utilizing the binary coding solved function obtained, and in Numerical Optimization, process is: first binary code is converted to decimal number, then substitutes in the function of optimization, obtains its functional value and is fitness.

Step 5: to each state computation fitness；

Step 6: if end condition is unsatisfactory for, then make t=t+1, then according to certain adjustable strategies, utilizes Quantum rotating gate operation that population at individual is updated, obtains progeny population Q (t+1), repeat step 3, step 4 and step 5；Updating by more having newly arrived to Quantum rotating gate of quantum bit, its process is:

$\left[\begin{array}{c}{\mathrm{\α}}_i^{\′}\\ {\mathrm{\β}}_i^{\′}\end{array}\right]=\left[\begin{array}{cc}\cos \left({\mathrm{\θ}}_i\right)& -\sin \left({\mathrm{\θ}}_i\right)\\ \sin \left({\mathrm{\θ}}_i\right)& \cos \left({\mathrm{\θ}}_i\right)\end{array}\right]\left[\begin{array}{c}{\mathrm{\α}}_i\\ {\mathrm{\β}}_i\end{array}\right]$

Wherein [α_i,β_i]^TIt is the quantum bit of i-th chromosome, θ_iFor the anglec of rotation；

Described adjustable strategies is as shown in the table:

Wherein x_iFor the i-th bit of current chromosome, BEST_iFor the current chromosomal i-th bit of optimum, f (x) is fitness function, Δ θ_iFor anglec of rotation size, the convergence rate of control algolithm, s (α_i,β_i) it is the direction of the anglec of rotation, it is ensured that convergence of algorithm, and meet θ_i=s (α_i,β_i)Δθ_i。

Step 7: record optimized individual and fitness, until end condition meets, stops algorithm output result.

Also utilize traditional genetic algorithm that above experimental data has been calculated, traditional genetic algorithm initial parameter such as following table used simultaneously:

Both result of calculations and calculating error contrast are as shown in the table:

Er is the calculating percentage error of each coordinate, and computing formula is as follows:

$Er=\left|\frac{L_{act}-L_{cal}}{L_{act}}\right|\×100\%$

Wherein L_actIt is actual coordinate value, L_calIt it is coordinates computed value.

D_maxFor maximum error value, computing formula is as follows:

$D_{max}=max\left\{\begin{array}{c}\left|x_{act}-x_{cal}\right|\\ \left|y_{act}-y_{cal}\right|\\ \left|z_{act}-z_{cal}\right|\end{array}\right\}$

Wherein x_act,y_act,z_actFor x, y, z actual coordinate value, x_cal,y_cal,z_calFor x, y, z coordinates computed value.

As can be seen from the above results compared with traditional genetic algorithm, quantum genetic algorithm has and under less population scale, can calculate iterations few, and positioning precision is higher, it is to avoid Premature Convergence and converge to the feature of globally optimal solution rapidly.

Above-listed detailed description is illustrating for possible embodiments of the present invention, and this embodiment is also not used to limit the scope of the claims of the present invention, and all equivalences done without departing from the present invention are implemented or change, and are intended to be limited solely by the scope of the claims of this case.

Claims (6)

1. the converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm, it is characterised in that it comprises the following steps:

Step one: adopt the sonac being placed in the multiple diverse location of transformator to receive the ultrasonic signal that Partial Discharge Sources sends, set up distance solving model with cartesian coordinate system；

Step 2: determine quantum genetic algorithm initial parameter, arranges maximum population G and population scale N from generation to generation, chromosome is encoded, makes t=0, initialize population Q (t)；

Step 3: to each individual enforcement one-shot measurement of initial population Q (t), obtain a state P (t)；

Step 4: this group solution of state P (t) carries out Fitness analysis, and record optimal adaptation degree individuality is as the desired value of next step evolution；

Step 5: each state computation fitness to initial population；

Step 6: if end condition is unsatisfactory for, then make t=t+1, then according to certain adjustable strategies, utilize Quantum rotating gate operation and quantum non-gate that population at individual is updated, obtain progeny population Q (t+1), repeat step 3, step 4 and step 5；

Step 7: record optimized individual and fitness, until end condition meets, stops algorithm output result.

2. the converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm according to claim 1, it is characterized in that, in described step one, the quantity of sonac is n, wherein n >=6, it is arranged on converter power transformer fuel tank outer wall, under fuel tank, cartesian coordinate system is set up for zero in summit, sets up nonlinear mathematical model:

$\left\{\begin{array}{c}\sqrt{{(x-x_2)}^2+{(y-y_2)}^2+{(z-z_2)}^2}-\sqrt{{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2}=v_s{\mathrm{\τ}}_{12}\\ \sqrt{{(x-x_3)}^2+{(y-y_3)}^2+{(z-z_3)}^2}-\sqrt{{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2}=v_s{\mathrm{\τ}}_{13}\\ ......\\ \sqrt{{(x-x_n)}^2+{(y-y_n)}^2+{(z-z_n)}^2}-\sqrt{{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2}=v_s{\mathrm{\τ}}_{1n}\end{array}\right.$

Set a trap to put and be a little positioned at that (x, y, in z), sensor location coordinates is followed successively by (x₁,y₁,z₁)、(x₂,y₂,z₂)……(x_n,y_n,z_n), equivalent velocity of sound v_sIf, with sensor (x₁,y₁,z₁) for trigger source, τ_1i, i=2,3 ..., n is each sensor and (x, y, the several signal time delay between z)；

If v_s∈ [1,1.6], (mm/ μ s), wherein mm/ μ s represents the every microsecond of millimeter；Depending on v_sThe unknown, in above formula, parameter to be asked is x, y, z, v_s, above-mentioned equation group is represented by:

f_i(x,y,z,v_s)=[(x-x_i)²+(y-y_i)²+(z-z_i)²]^1/2-[(x-x₁)²+(y-y₁)²+(z-z₁)²]^1/2-v_sτ_1i=0

Adding constraints, above formula becomes:

$\begin{array}{c}\min \left\{D_f(x,y,z,v_s)\right\}=\\ {\mathrm{\Σ}}_{i=2}^n{\{{\[{(x-x_i)}^2+{(y-y_i)}^2+{(z-z_i)}^2\]}^{1/2}+{\[{(x-x_1)}^2+{(y-y_1)}^2+{(z-z_1)}^2\]}^{1/2}-v_s{\mathrm{\τ}}_{1i}\}}^2\end{array}$

Constraints:

$\left.\begin{array}{c}0\≤x\≤x_{\max }\\ 0\≤y\≤y_{\max }\\ 0\≤z\≤z_{\max }\\ 1\≤v_s\≤1.6,(mm/\mathrm{\μ}s)\end{array}\right\}$

The length coordinate of a position, v are put in x, y, z respectively office_sFor converter power transformer internal ultrasonic ripple equivalence velocity of wave, x_max,y_max,z_maxFor the actual length of converter power transformer tank wall；

Above formula is transcendental equations, and when obtaining physical constraint, optimal solution obtains optimal solution when physical constraint.

3. the converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm according to claim 1, it is characterised in that in described step 2, chromosome is carried out quantum bit coding, each quantum individuality is with following quantum form coding:

$q_j^t=\[\left.\begin{array}{c}{\mathrm{\α}}_{11}^t\\ {\mathrm{\β}}_{11}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{12}^t\\ {\mathrm{\β}}_{12}^t\end{array}\right|\left.\begin{array}{c}...\\ ...\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{1k}^t\\ {\mathrm{\β}}_{1k}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{21}^t\\ {\mathrm{\β}}_{21}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{22}^t\\ {\mathrm{\β}}_{22}^t\end{array}\right|\left.\begin{array}{c}...\\ ...\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{2k}^t\\ {\mathrm{\β}}_{2k}^t\end{array}\right|\left.\begin{array}{c}...\\ ...\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{n1}^t\\ {\mathrm{\β}}_{n1}^t\end{array}\right|\left.\begin{array}{c}{\mathrm{\α}}_{n2}^t\\ {\mathrm{\β}}_{n2}^t\end{array}\right|\left.\begin{array}{c}...\\ ...\end{array}\right|\begin{array}{c}{\mathrm{\α}}_{nk}^t\\ {\mathrm{\β}}_{nk}^t\end{array}\]$

WhereinBeing that t is individual for jth, n is the quantum gene number that each quantum is individual, and k is the quantum bit number used by the component of each independent variable, α_ij,β_ijIt is that two plural numbers are called probability amplitude pair, and meets: | α_ij|²+|β_ij|²=1, | α_ij|²For quantum be in spin downward probability of state, | β_ij|²It is in for quantum and spins up probability of state；

In described initialization of population, population scale is N, namely has N number of quantum coding individual, and each quantum individuality is set toNamely

$\left[\begin{array}{c}\frac{1}{\sqrt{2}}\frac{1}{\sqrt{2}}...\frac{1}{\sqrt{2}}\\ \frac{1}{\sqrt{2}}\frac{1}{\sqrt{2}}...\frac{1}{\sqrt{2}}\end{array}\right].$

4. the converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm according to claim 1, it is characterised in that during measurement, according to quantum bit probability amplitude | α_ij|²Or | β_ij|²Select 0 on corresponding gene position or 1, method particularly includes: randomly generate one [0,1], if it is be more than or equal to probability amplitude | α_ij|²Or | β_ij|²Value, then measurement result takes 1；Otherwise taking 0, vice versa, thus the individuality of quantum coding is converted to binary-coded individuality, obtains N number of binary-coded individuality.

5. the converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm according to claim 1, it is characterized in that, in described step 4, the described fitness that solves refers to the fitness utilizing the binary coding solved function obtained, in Numerical Optimization, process is: first binary code is converted to decimal number, then substitute in function to be optimized, obtain its functional value and be fitness.

6. the converter power transformer type local-discharge ultrasonic localization method based on quantum genetic algorithm according to claim 1, it is characterised in that in described step 6, the renewal of quantum bit is by more having newly arrived to quantum door, and its process is:

$\left[\begin{array}{c}{\mathrm{\α}}_i^{\′}\\ {\mathrm{\β}}_i^{\′}\end{array}\right]=\left[\begin{array}{cc}\cos \left({\mathrm{\θ}}_i\right)& -\sin \left({\mathrm{\θ}}_i\right)\\ \sin \left({\mathrm{\θ}}_i\right)& \cos \left({\mathrm{\θ}}_i\right)\end{array}\right]\left[\begin{array}{c}{\mathrm{\α}}_i\\ {\mathrm{\β}}_i\end{array}\right]$

Wherein [α_i,β_i]^TIt is the quantum bit of i-th chromosome, θ_iFor the anglec of rotation；

Described adjustable strategies is as shown in the table:

Wherein x_iFor the i-th bit of current chromosome, BEST_iFor the current chromosomal i-th bit of optimum, f (x) is fitness function, Δ θ_iFor anglec of rotation size, the convergence rate of control algolithm, s (α_i,β_i) it is the direction of the anglec of rotation, it is ensured that convergence of algorithm, and meet θ_i=s (α_i,β_i)Δθ_i。