CN106771847A - A kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method - Google Patents

Abstract

The present invention provides a kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method, including：Transmission line lightning stroke risk parameter measurement is carried out to 35kV power distribution networks；Measurement data normalized；The thunderbolt risk of subsequent time is predicted for the measurement data after normalized, the more big possibility that thunderbolt risk then occurs of 35kV power distribution network transmission line lightning strokes risk index is bigger.The present invention carries out real-time monitoring for power distribution network operational factor and geographical environment parament, chooses power distribution network operational factor and place environmental geography environment parament；And the prediction of 35kV power distribution network transmission line lightning strokes risk index is predicted according to monitoring parameter, power distribution network is controlled in real time according to predicting the outcome, loss of the power distribution network caused by the accidents such as thunderbolt power failure is prevented effectively from, the reliability and economy of power distribution network Operation of Electric Systems is significantly improved.

Description

A kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method

Technical field

The invention belongs to power distribution network technical field of electric power transmission, more particularly to a kind of 35kV power distribution networks transmission line lightning stroke risk is pre- Survey method.

Background technology

In 35kV power distribution networks, thunder and lightning hits transmission line of electricity can cause line tripping, load to have a power failure, to power system and negative Lotus user equipment is likely to bring substantial equipment to lose and casualties.How according to distribution network structure structure, geographical meteorology Condition, power network and load operating conditions, are effectively predicted 35kV power distribution network transmission line lightning stroke indexes, circuit is struck by lightning and is jumped Lock risk makes assessment, realize 35kV power distribution network transmission line lightning strokes risk index prediction can ensure power distribution network it is safe and stable, Effec-tive Function.The characteristics of conventional 35kV power distribution network transmission line lightning stroke Risk Calculation methods is to ignore grid structure with geography meteorology Interaction relationship between condition, is calculated transmission line lightning stroke possibility size, it is impossible to have by single meteorological condition Effect is using operation of power networks status data and geographical meteorology integrated data resource, and accuracy in computation and reliability be not high.

The content of the invention

In view of the shortcomings of the prior art, the present invention provides a kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method.

Technical scheme is as follows：

A kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method, including：

Step 1：Transmission line lightning stroke risk parameter measurement is carried out to 35kV power distribution networks；The transmission line lightning stroke risk ginseng Number, including：Temperature, humidity, the air pressure of environment where transmission line of electricity voltage, transmission line of electricity electric current, 35kV power distribution networks；

Step 2：Measurement data normalized；

Step 3：Using 35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings, for the survey after normalized Amount data are predicted to the thunderbolt risk of subsequent time；The 35kV power distribution networks transmission line lightning stroke risk index Mathematical Modeling Input be the measurement data after normalized, be output as 35kV power distribution network transmission line lightning stroke risk indexs, 35kV distribution The more big possibility that thunderbolt risk then occurs of net transmission line lightning stroke risk index is bigger.

The method of measurement data normalized is as follows in the step 2：

Following normalized is carried out for measurement data：

Wherein, i=1,2 ..., 5n, n represent the number to every class measurement data acquisition, glx_iRepresent that i-th normalization is surveyed Amount data, rglx_iRepresent ith measurement data, rglxi_max、rglx_iminThe maximum and most of the i-th class measurement data is represented respectively Small value.

The step 3, including：

Step 3.1：Set up the 35kV power distribution network transmission line lightning stroke risk index mathematics with penalty factor and bound term Model：

y_ayz=minf_mb(glx_i)+g_cf(glx_i)+r_ys(glx_i)

Wherein, y_ayzIt is 35kV power distribution network transmission line lightning stroke risk indexs, object function35kV power distribution network transmission line lightning stroke risk index object functions Penalty factor g_ef(glx_i)=ln (max (glx_i)-min(glx_i)), 35kV power distribution network transmission line lightning stroke risk index target letters Several bound termmax(glx_i) it is the maximum in the measurement data after normalized Value, min (glx_i) it is the minimum value in the measurement data after normalized；

Step 3.2：35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings are solved using SVMs, is obtained 35kV power distribution network transmission line lightning stroke risk index predicted values.

The step 3.2, including：

Step 3.2.1：Choose the kernel function of SVMsWherein, | glx_i+1-glx_i| it is glx_i+1With glx_iBetween distance, σ be not equal to zero constant；

Step 3.2.2：Using the σ in the kernel function of artificial immune network Training Support Vector Machines；

Step 3.2.3：35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings are solved, 35kV power distribution networks is obtained defeated Electric line thunderbolt risk exponential forecasting value.

The step 3.2.2, including：

Step 3.2.2.1：By target function value A_psAs the prototype that artificial immune network is cloned, it is determined that the number of clone N_clo=K_CCA_ps, wherein, N_cloIt is the number of clone, K_CCIt is clone's constant；

Step 3.2.2.2：Number according to clone obtains cloning matrixIt is thereinBe from Memory in randomly selected measurement data in measurement data after all normalizeds, i.e. original manual immunological network resists Body；

Step 3.2.2.3：To clone's matrixEnter row variation to obtain cloning result：

C_clo(X_pps)=C_clo(X_pps)-α_pps(C_clo(X_pps)-glx_i)

Wherein, X_pps(X_pps=1,2 ..., Nx_lo) for the aberration rate α of clone_pps=rand () × | | C_clo(X_pps)- glx_i||×R_pps, R_ppsIt is variation constant, rand () is the random number between [0,1]；

Step 3.2.2.4：TraversalIn all of memory antibody, perform the mutation operation of step 3.2.2.3；

Step 3.2.2.5：Determine memory antibodyTwo minimum memory antibodies of middle difference, by it from clone Matrix C_cloMiddle deletion, the generation for so far completing artificial immune network is evolved；

Step 3.2.2.6：Terminate to evolve when evolutionary generation reaches the evolutionary generation genc of setting, try to achieve| | | | it is vector norm.

Beneficial effect：

The present invention carries out real-time monitoring for power distribution network operational factor and geographical environment parament, chooses distribution network operation Environmental geography environment parament where parameter -- transmission line of electricity voltage, transmission line of electricity electric current, and 35kV power distribution networks --- temperature Degree, humidity, air pressure；And the prediction of 35kV power distribution network transmission line lightning strokes risk index is predicted according to monitoring parameter, according to Result of calculation is controlled to power distribution network in real time, can be prevented effectively from damage of the distribution network system caused by the accidents such as thunderbolt power failure Lose, significantly improve the reliability and economy of power distribution network Operation of Electric Systems.

Brief description of the drawings

Fig. 1 is the 35kV power distribution network transmission line lightning stroke Risk Forecast Method flow charts in the specific embodiment of the invention；

Fig. 2 is the flow chart of the step 3 in the specific embodiment of the invention；

Fig. 3 is the flow chart of the step 3.2.2 in the specific embodiment of the invention.

Specific embodiment

Specific embodiment of the invention is elaborated below in conjunction with the accompanying drawings.

A kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method, as shown in figure 1, including：

Step 1：Transmission line lightning stroke risk parameter measurement is carried out to 35kV power distribution networks；The transmission line lightning stroke risk ginseng Number, including：Temperature, humidity, the air pressure of environment where transmission line of electricity voltage, transmission line of electricity electric current, 35kV power distribution networks；

According to the measurement data of collection, following sequence is set up according to sampling instant：

In Fixed Time Interval to transmission line of electricity voltage, transmission line of electricity electric current, temperature, humidity, air pressure are measured, then existed A series of sampling instant tlt₁, tlt₂..., tlt_n(n is natural number, n=1,2 ...) obtains measured value：Transmission line of electricity voltage tslU₁, tslU₂..., tslU_n, transmission line of electricity electric current tslI₁, tslI₂..., tslI_n, temperature tszt₁, tszt₂..., tszt_n, humidity tszh₁, tszh₂..., tszh_n, air pressure tszp₁, tszp₂..., tszp_n。

Step 2：It is uniform data dimension and excursion, to measurement data normalized；

Following normalized is carried out for measurement data：

Wherein, i=1,2 ..., 5n, n represent the number to every class measurement data acquisition, glx_iRepresent that i-th normalization is surveyed Amount data, rglx_iRepresent ith measurement data, rglx_imax、rglx_iminThe maximum and most of the i-th class measurement data is represented respectively Small value.

Step 3：Using 35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings, for the survey after normalized Amount data are predicted to the thunderbolt risk of subsequent time；The 35kV power distribution networks transmission line lightning stroke risk index Mathematical Modeling Input be the measurement data after normalized, be output as 35kV power distribution network transmission line lightning stroke risk indexs, 35kV distribution The more big possibility that thunderbolt risk then occurs of net transmission line lightning stroke risk index is bigger.

The step 3, as shown in Fig. 2 including：

Step 3.1：Set up the 35kV power distribution network transmission line lightning stroke risk index mathematics with penalty factor and bound term Model：

y_ayz=minf_mb(glx_i)+g_cf(glx_i)+r_ys(glx_i)

Wherein, y_ayzIt is 35kV power distribution network transmission line lightning stroke risk indexs, object function35kV power distribution network transmission line lightning stroke risk index object functions Penalty factor g_ef(glx_i)=ln (max (glx_i)-min(glx_i)), 35kV power distribution network transmission line lightning stroke risk index target letters Several bound termmax(glx_i) it is the maximum in the measurement data after normalized Value, min (glx_i) it is the minimum value in the measurement data after normalized；

Step 3.2：35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings are solved using SVMs, is obtained 35kV power distribution network transmission line lightning stroke risk index predicted values.

The step 3.2, including：

Step 3.2.1：Choose the kernel function of SVMsWherein, | glx_i+1-glx_i| it is glx_i+1With glx_iBetween distance, σ be not equal to zero constant；

Step 3.2.2：Using the σ in the kernel function of artificial immune network Training Support Vector Machines；

Specific steps as shown in figure 3, including：

Step 3.2.2.1：By target function value A_psAs the prototype that artificial immune network is cloned, it is determined that the number of clone N_clo=K_CCA_ps, wherein, N_cloIt is the number of clone, K_CCIt is clone's constant, takes K_CC=1.057,；

Step 3.2.2.2：Number according to clone obtains cloning matrixIt is thereinBe from Memory in randomly selected measurement data in measurement data after all normalizeds, i.e. original manual immunological network resists Body；

Step 3.2.2.3：To clone's matrixEnter row variation to obtain cloning result：

C_clo(X_pps)=C_clo(X_pps)-α_pps(C_clo(X_pps)-glx_i)

Wherein, X_pps(X_pps=1,2 ..., N_clo) for the aberration rate α of clone_pps=rand () × | | C_clo(X_pps)-glx_i ||×R_pps, R_ppsIt is variation constant, takes R_pps=0.8253, rand () are the random number between [0,1]；

Step 3.2.2.4：TraversalIn all of memory antibody, perform the mutation operation of step 3.2.2.3；

Step 3.2.2.5：Determine memory antibodyTwo minimum memory antibodies of middle difference, by it from clone Matrix C_cloMiddle deletion, the generation for so far completing artificial immune network is evolved；

Step 3.2.2.6：Terminate to evolve when evolutionary generation reaches the evolutionary generation genc=13 of setting, try to achieveIt is vector norm.

Step 3.2.3：35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings are solved, 35kV power distribution networks is obtained defeated Electric line thunderbolt risk exponential forecasting value.

Claims (5)

1. a kind of 35kV power distribution networks transmission line lightning stroke Risk Forecast Method, it is characterised in that including：

Step 1：Transmission line lightning stroke risk parameter measurement is carried out to 35kV power distribution networks；The transmission line lightning stroke risk parameter, Including：Temperature, humidity, the air pressure of environment where transmission line of electricity voltage, transmission line of electricity electric current, 35kV power distribution networks；

Step 2：Measurement data normalized；

Step 3：Using 35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings, for the measurement number after normalized It is predicted according to the thunderbolt risk to subsequent time；The 35kV power distribution networks transmission line lightning stroke risk index Mathematical Modeling it is defeated Measurement data after entering for normalized, is output as 35kV power distribution network transmission line lightning stroke risk indexs, and 35kV power distribution networks are defeated The more big possibility that thunderbolt risk then occurs of electric line thunderbolt risk index is bigger.

2. method according to claim 1, it is characterised in that the method for measurement data normalized in the step 2 It is as follows：

Following normalized is carried out for measurement data：

${\mathrm{glx}}_i=\frac{{\mathrm{rglx}}_i-{\mathrm{rglx}}_{min}}{{\mathrm{rglx}}_{\max }-{\mathrm{rglx}}_{min}}$

Wherein, i=1,2 ..., 5n, n represent the number to every class measurement data acquisition, glx_iRepresent i-th AVHRR NDVI number According to rglx_iRepresent ith measurement data, rglx_imax、rglx_iminThe maximum and minimum of the i-th class measurement data are represented respectively Value.

3. method according to claim 1, it is characterised in that the step 3, including：

Step 3.1：Set up the 35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings with penalty factor and bound term：

y_ayz=min f_mb(glx_i)+g_cf(glx_i)+r_ys(glx_i)

Wherein, y_ayzIt is 35kV power distribution network transmission line lightning stroke risk indexs, object function The penalty factor g of 35kV power distribution network transmission line lightning stroke risk index object functions_ef(glx_i)=ln (max (glx_i)-min (glx_i)), the bound term of 35kV power distribution network transmission line lightning stroke risk index object functions max(glx_i) it is the maximum in the measurement data after normalized, min (glx_i) it is the measurement data after normalized In minimum value；

Step 3.2：35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings are solved using SVMs, 35kV is obtained Power distribution network transmission line lightning stroke risk index predicted value.

4. method according to claim 3, it is characterised in that the step 3.2, including：

Step 3.2.1：Choose the kernel function of SVMsWherein, | glx_i+1-glx_i | it is glx_i+1With glx_iBetween distance, σ be not equal to zero constant；

Step 3.2.2：Using the σ in the kernel function of artificial immune network Training Support Vector Machines；

Step 3.2.3：35kV power distribution network transmission line lightning stroke risk index Mathematical Modelings are solved, 35kV power distribution network power transmission lines are obtained Road thunderbolt risk exponential forecasting value.

5. method according to claim 4, it is characterised in that the step 3.2.2, including：

Step 3.2.2.1：By target function value A_psAs the prototype that artificial immune network is cloned, it is determined that the number N of clone_clo= K_CCA_ps, wherein, N_cloIt is the number of clone, K_CCIt is clone's constant；

Step 3.2.2.2：Number according to clone obtains cloning matrixIt is thereinIt is from all Memory antibody in randomly selected measurement data in measurement data after normalized, i.e. original manual immunological network；

Step 3.2.2.3：To clone's matrixEnter row variation to obtain cloning result：

C_clo(X_pps)=C_clo(X_pps)-α_pps(C_clo(X_pps)-glx_i)

Wherein, X_pps(X_pps=1,2 ..., N_clo) for the aberration rate α of clone_pps=rand () × | | C_clo(X_pps)-glx_i|| ×R_pps, R_ppsIt is variation constant, rand () is the random number between [0,1]；

Step 3.2.2.4：TraversalIn all of memory antibody, perform the mutation operation of step 3.2.2.3；

Step 3.2.2.5：Determine memory antibodyTwo minimum memory antibodies of middle difference, by it from clone's matrix C_cloMiddle deletion, the generation for so far completing artificial immune network is evolved；

Step 3.2.2.6：Terminate to evolve when evolutionary generation reaches the evolutionary generation genc of setting, try to achieve| | | | it is vector norm.