A method for predicting ice dancing of transmission lines

Abstract

The present invention relates to the technical field of power transmission lines, and in particular to a method for predicting ice dancing of power transmission lines. The present invention adopts a big data statistical method, and for different icing types, selects the transmission line with the highest prediction accuracy as the representative transmission line; for different icing types, calls the transmission line ice dancing prediction model with the highest prediction accuracy to complete the ice dancing prediction of the representative transmission line; for the remaining transmission lines, the ice dancing prediction accuracy is low, and whether ice dancing will occur is judged based on the linkage probability with the representative transmission line. With this technical solution, the prediction result of the representative transmission line is obtained through big data analysis, proving that the icing form and wind direction angle of the representative transmission line are prone to ice dancing with a high probability, and it is no longer necessary to quantify the icing form, and the remaining transmission lines are quantified according to the difference between them and the representative transmission line, so as to complete the probability prediction of the associated dancing.

Description

Technical Field

[0001] The present invention relates to the technical field of power transmission lines, and in particular to a method for predicting ice dancing of power transmission lines. Background Art

[0002] Transmission lines are the core foundation of the national power grid. In cold environments, transmission lines are very prone to ice accumulation, which may cause accidents such as line breakage and ice dancing, especially ice dancing, which is more likely to occur.

[0003] The causes of icing are affected by atmospheric temperature, humidity, wind speed, wind direction, cold and warm air currents, etc. By analyzing climate factors, early warning of icing can be given to a certain extent. Through in-depth research, taking formation mechanism and adhesion ability as considerations, icing can be divided into five categories, namely rain rime, fog rime, mixed rime, dry snow rime, and wet snow rime.

[0004] The galloping of ice-covered transmission lines is caused by aerodynamic instability caused by irregular ice. The biggest influencing factors are the ice quality, ice form, geological environment, wind direction and wind speed of the transmission line, among which the ice quality and ice form cannot be quantified. Therefore, how to complete the ice galloping prediction of transmission lines in the absence of ice quality and ice form is an urgent problem to be solved. Summary of the invention

[0005] The present invention discloses a method for predicting ice dancing of a power transmission line. The specific method is as follows:

[0006] Obtain current climate information at the time of diagnosis to determine possible ice cover types;

[0007] According to the possible icing type information, select the representative transmission line number in the diagnosis area as a representative;

[0008] According to the representative transmission line number, the corresponding ice dancing prediction model is retrieved;

[0009] Input the ice dancing prediction model with current climate information to obtain the ice dancing prediction results of representative transmission lines;

[0010] According to the linkage ice dancing probability scores of the representative transmission line and the other transmission lines, combined with the ice dancing prediction results of the representative transmission line, the ice dancing probability of the other transmission lines in the diagnosis area is calculated.

[0011] The advantage of this embodiment is that, by adopting the method of big data statistics, for frost, rime, mixed frost, dry snow and frost, and wet snow and frost, the transmission line with the highest prediction accuracy is selected as the representative transmission line; when encountering a certain icing climate, the transmission line icing dance prediction model with the highest prediction accuracy is called to complete the icing dance prediction of the representative transmission line; for the remaining transmission lines, the icing dance prediction accuracy is low, and the icing dance is not judged by the model prediction method, but instead judged whether the icing dance will occur based on the linkage probability with the representative transmission line. With this technical solution, the prediction result of the representative transmission line is obtained through big data analysis, proving that the icing form and wind direction angle of the representative transmission line are prone to icing dance with a high probability, and there is no need to quantify the icing form. The remaining transmission lines are quantified according to their differences with the representative transmission line, and the probability prediction of the associated dance can be completed, and there is no need to quantify the icing form, and the number of large models used is significantly reduced, saving computing power.

[0012] Furthermore, the current climate information includes: temperature, humidity, wind speed and wind direction.

[0013] Furthermore, the ice cover types include: rain rime, fog rime, mixed rime, dry snow rime and wet snow rime.

[0014] Furthermore, a representative transmission line number is selected as a representative in the diagnosis area, and the specific method is as follows:

[0015] Mark all transmission lines in the diagnosis area with unique numbers;

[0016] Retrieve historical information of the diagnosis area;

[0017] In the statistical historical information, the transmission line with the most ice dancing times in each ice type is taken as the representative transmission line.

[0018] The advantage of this embodiment is that by counting historical information, the representative transmission line that can best predict each type of ice cover can be found. The climate of a region changes greatly in a short period of time, and a single transmission line cannot be accurately judged. However, big data statistics can eliminate short-term uncertainties and find the transmission line with the smallest change as the representative transmission line. The prediction accuracy for the representative transmission line can be significantly improved.

[0019] Furthermore, the ice dancing prediction model corresponds one-to-one to the representative transmission line and is an LSTM neural network model.

[0020] Furthermore, the linkage ice dancing probability score between the transmission line and other transmission lines is obtained as follows:

[0021] According to the historical information under different climate information, the coefficient of difference in unevenness, the coefficient of difference in average ice thickness, the coefficient of difference in wind direction and the coefficient of difference in wind speed between the representative transmission line and the comparison transmission line are calculated;

[0022] The weighted unevenness difference coefficient, the average ice thickness difference coefficient, the wind direction difference coefficient and the wind speed difference coefficient are used as the linkage ice dancing probability scores for the representative transmission line and the rest of the transmission lines.

[0023] The advantage of this embodiment is that when predicting whether ice dancing will occur, the current unevenness and ice thickness of the transmission line cannot be measured in real time due to cost considerations; however, under specific ice types, the difference from the representative transmission line can be obtained through data statistics. Therefore, as long as there is no change in the macroclimatic environment, this difference conforms to the normal distribution. Replacing real-time measurement with historical anomalies can reflect the current state of the transmission line as accurately as possible while significantly saving costs, thereby obtaining the probability of associated dancing of the remaining transmission lines and the representative transmission line.

[0024] Furthermore, the coefficient of unevenness is calculated as follows:

[0025]

[0026] in

[0027]

[0028] Where, d u is the coefficient of unevenness, u rep represents the uneven ice coverage of the transmission line, u rem For the uneven ice coverage of other power lines, represents the length of the transmission line where the ice thickness exceeds the preset value. represents the length of the transmission line where the ice thickness is less than the preset value. For the remaining transmission lines where ice thickness exceeds the preset value, The length of the remaining transmission lines where the ice thickness is less than the preset value;

[0029] The coefficient of variation of average ice thickness is calculated as follows:

[0030]

[0031] in

[0032]

[0033] Where, d tk is the average ice thickness variation coefficient, tk rep represents the unit ice mass of the transmission line, tkrem is the unit ice mass of other power lines, m rep represents the quality of the transmission line, L rep represents the length of the transmission line, m rem is the quality of the remaining transmission lines, L rem is the remaining transmission line length.

[0034] Furthermore, the wind direction difference coefficient is calculated as follows:

[0035]

[0036] Where, d θ is the wind direction difference coefficient, n is the historical information number, is the wind direction angle representing the transmission line in the nth history, is the wind direction angle of the remaining transmission lines in the nth history;

[0037] Wind speed difference coefficient, calculated as follows:

[0038]

[0039] Where, d v is the wind speed difference coefficient, n is the historical information number, is the wind speed of the transmission line in the nth history, is the wind speed of the remaining transmission lines in the nth history.

[0040] The advantage of this embodiment is that the unevenness difference coefficient, the ice average thickness difference coefficient, the wind direction difference coefficient and the wind speed difference coefficient are customized, which provides a basis for calculating the associated ice dancing.

[0041] Furthermore, the probability score of ice dancing is linked, and the calculation formula is as follows:

[0042] s=ω1.d u +ω2.d tk +ω3.d θ +ω4.d v

[0043] Where ω1, ω2, ω3 and ω4 are weights, which are dynamically adjusted in real time according to the diagnosis accuracy of the remaining transmission lines.

[0044] According to the diagnosis of ice dancing prediction model, the probability of ice dancing on the transmission line is p rep ;

[0045] Then the probability of ice dancing on the remaining transmission lines is p rem =(1-s)×p rep .

[0046] Furthermore, the specific method of real-time dynamic adjustment is as follows:

[0047] Calculate the prediction results of the remaining transmission lines;

[0048] If the probability of linked ice dancing between the remaining transmission lines and the representative transmission line is less than a preset value, the difference in unevenness, average ice thickness, wind direction and wind speed between the representative transmission line and the remaining transmission lines is analyzed;

[0049] Select the attribute with the highest difference ranking as the attribute to be adjusted;

[0050] Based on the weight corresponding to the attribute to be adjusted, the weight corresponding to the attribute to be adjusted is reduced according to the coefficient ranked first in the preset proportional order;

[0051] Recalculate the newly generated prediction results of the transmission line. If the probability of ice dancing between the transmission line and the representative transmission line is greater than the preset value, stop adjusting the weight.

[0052] If the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value, and the attribute with the highest difference ranking has not changed, the weight corresponding to the attribute to be adjusted will be reduced according to the coefficient ranked last in the preset proportional order; if the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value, and the attribute with the highest difference ranking has changed, the weight corresponding to the attribute to be adjusted will be changed, and the weight corresponding to the attribute to be adjusted will be reduced according to the coefficient ranked first in the preset proportional order until the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value.

[0053] The advantage of this embodiment is that when some difference coefficients are too large, the influence of the difference coefficients can be reduced by reducing the weight ratio, thereby playing a role in real-time adjustment and optimization; this embodiment proposes a rotational adjustment method for proportionally reducing the weights, which can timely and accurately complete the correction of the associated dance calculation formula, and significantly improve the prediction accuracy.

[0054] Other advantages, objectives and features of the present invention will be described in the following description to some extent, and to some extent, will be obvious to those skilled in the art based on the following examination and study, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings of the present invention are as follows.

[0056] Figure 1 This is a diagram of the ice coverage status of transmission lines.

[0057] Figure 2 It is a schematic diagram of the overall process of the present invention. DETAILED DESCRIPTION

[0058] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0059] A method for predicting ice dancing of transmission lines, such as Figure 2 As shown, the specific steps are as follows:

[0060] S1. Obtain current climate information during diagnosis and determine possible ice cover types.

[0061] In this embodiment, the current climate information includes: temperature, humidity, wind speed and wind direction. According to the changes in temperature, humidity, wind speed and wind direction, the possibility and type of icing on the transmission line can be roughly determined. The types of icing include: rain rime, fog rime, mixed rime, dry snow rime and wet snow rime. The icing status of the transmission line is as follows: Figure 1 shown.

[0062] A description of each icing type is shown in the table below.

[0063]

[0064] As can be seen from the table above, the mass density and adhesion of each type of ice are different, and the ice morphology is also different. Combined with different wind speeds and wind directions, it is difficult to determine whether ice dancing will occur. Therefore, big data statistics can be used to find one of the representative transmission lines as the prediction target.

[0065] S2. Select a representative transmission line number in the diagnosis area as a representative according to the possible icing type information.

[0066] Select the representative transmission line number in the diagnosis area as a representative. The specific steps are as follows:

[0067] S21. Mark all power transmission lines in the diagnosis area with unique numbers.

[0068] S22. Retrieve historical information of the diagnosis area.

[0069] S23. In the statistical historical information, the transmission line with the most ice dancing times in each ice type is taken as the representative transmission line.

[0070] In this embodiment, the representative transmission line selected for the rain and frost icing type is N1; the representative transmission line selected for the fog and frost icing type is N2; the representative transmission line selected for the mixed frost icing type is N3; the representative transmission line selected for the dry snow and frost icing type is N4; and the representative transmission line selected for the wet snow and frost icing type is N5.

[0071] S3. According to the representative transmission line number, the corresponding ice dancing prediction model is retrieved.

[0072] In this embodiment, taking the frost and ice type as an example, the representative transmission line selected is N1, and the neural network model is f LSTM (N1).

[0073] S4. Input the current climate information into the ice dancing prediction model to obtain the ice dancing prediction result of the representative transmission line.

[0074] In this embodiment, the current temperature, humidity, wind speed and wind direction are input into f LSTM In (N1), f LSTM (N1) output represents the probability p of ice dancing of transmission line N1 rep .

[0075] S5. Calculate the probability of ice dancing of the remaining transmission lines in the diagnosis area based on the linkage ice dancing probability scores of the representative transmission line and the remaining transmission lines and the ice dancing prediction results of the representative transmission line.

[0076] In this embodiment, the remaining transmission lines also include the remaining transmission lines N2, N3, N4..., N i , for the remaining transmission lines N i , retrieve the ice dancing probability score s corresponding to the representative transmission line N1 1,i .

[0077] Combined with formula p rem =(1-s 1,i )×p rep , calculate the remaining transmission lines N i The probability of ice dancing.

[0078] Specifically, the linkage ice dancing probability score between the transmission line and other transmission lines is obtained as follows:

[0079] According to the historical information under different climate information, the coefficient of difference in unevenness, the coefficient of difference in average ice thickness, the coefficient of difference in wind direction and the coefficient of difference in wind speed between the representative transmission line and the comparison transmission line are calculated;

[0080] The weighted unevenness difference coefficient, the average ice thickness difference coefficient, the wind direction difference coefficient and the wind speed difference coefficient are used as the linkage ice dancing probability scores for the representative transmission line and the rest of the transmission lines.

[0081] The coefficient of unevenness is calculated as follows:

[0082]

[0083] in

[0084]

[0085] Where, du is the coefficient of unevenness, u rep represents the uneven ice coverage of the transmission line, u rem For the uneven ice coverage of other power lines, represents the length of the transmission line where the ice thickness exceeds the preset value. represents the length of the transmission line where the ice thickness is less than the preset value. For the remaining transmission lines where ice thickness exceeds the preset value, The length of the remaining transmission lines where the ice thickness is less than the preset value;

[0086] The coefficient of variation of average ice thickness is calculated as follows:

[0087]

[0088] in

[0089]

[0090] Where, d tk is the average ice thickness variation coefficient, tk rep represents the unit ice mass of the transmission line, tk rem is the unit ice mass of other power lines, m rep represents the quality of the transmission line, L rep represents the length of the transmission line, m rem is the quality of the remaining transmission lines, L rem is the remaining transmission line length.

[0091] Wind direction difference coefficient, calculated as follows:

[0092]

[0093] Where, d θ is the wind direction difference coefficient, n is the historical information number, is the wind direction angle representing the transmission line in the nth history, is the wind direction angle of the remaining transmission lines in the nth history;

[0094] Wind speed difference coefficient, calculated as follows:

[0095]

[0096] Where, d v is the wind speed difference coefficient, n is the historical information number, is the wind speed of the transmission line in the nth history, is the wind speed of the remaining transmission lines in the nth history.

[0097] Linked ice dancing probability score, the calculation formula is as follows:

[0098] s=ω1.d u +ω2.d tk +ω3.d θ +ω4.d v

[0099] Where ω1, ω2, ω3 and ω4 are weights, which are dynamically adjusted in real time according to the diagnosis accuracy of the remaining transmission lines.

[0100] According to the diagnosis of ice dancing prediction model, the probability of ice dancing on the transmission line is p rep ;

[0101] Then the probability of ice dancing on the remaining transmission lines is p rem =(1-s)×p rep .

[0102] In this step, the specific steps of real-time dynamic adjustment are as follows:

[0103] S51, calculating the prediction results of the remaining transmission lines;

[0104] S52, if the probability of linked ice dancing between the remaining transmission lines and the representative transmission line is less than a preset value, analyzing the difference in unevenness, average ice thickness, wind direction, and wind speed between the representative transmission line and the remaining transmission lines;

[0105] S53, selecting the attribute with the highest difference ranking as the attribute to be adjusted;

[0106] S54, based on the weight corresponding to the attribute to be adjusted, reducing the weight corresponding to the attribute to be adjusted according to the coefficient ranked first in the preset proportional order;

[0107] S55, re-counting the newly generated prediction results of the transmission line, if the probability of ice dancing between the transmission line and the representative transmission line is greater than a preset value, then stop adjusting the weight;

[0108] If the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value, and the attribute with the highest difference ranking has not changed, the weight corresponding to the attribute to be adjusted will be reduced according to the coefficient ranked last in the preset proportional order; if the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value, and the attribute with the highest difference ranking has changed, the weight corresponding to the attribute to be adjusted will be changed, and the weight corresponding to the attribute to be adjusted will be reduced according to the coefficient ranked first in the preset proportional order until the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A method for predicting ice dancing of transmission lines, characterized in that: The specific method is as follows: Obtain current climate information at the time of diagnosis to determine possible ice cover types; According to the possible icing type information, select the representative transmission line number in the diagnosis area as a representative; According to the representative transmission line number, the corresponding ice dancing prediction model is retrieved; Input the ice dancing prediction model with current climate information to obtain the ice dancing prediction results of representative transmission lines; According to the linkage ice dancing probability scores of the representative transmission line and the other transmission lines, combined with the ice dancing prediction results of the representative transmission line, the ice dancing probability of the other transmission lines in the diagnosis area is calculated.

2. The method for predicting ice dancing of transmission lines according to claim 1, characterized in that: The current climate information includes: temperature, humidity, wind speed and wind direction.

3. The method for predicting ice dancing of transmission lines according to claim 1, characterized in that: The ice cover types include: rain rime, fog rime, mixed rime, dry snow rime and wet snow rime.

4. The method for predicting ice dancing of transmission lines according to claim 1, characterized in that: Select the representative transmission line number as the representative in the diagnosis area, the specific method is as follows: Mark all transmission lines in the diagnosis area with unique numbers; Retrieve historical information of the diagnosis area; In the statistical historical information, the transmission line with the most ice dancing times in each ice type is taken as the representative transmission line.

5. The method for predicting ice dancing of transmission lines according to claim 1, characterized in that: The ice dancing prediction model corresponds one-to-one to the representative transmission line and is an LSTM neural network model.

6. The method for predicting ice dancing of transmission lines according to claim 1, characterized in that: Represents the linkage ice dancing probability score of the transmission line and other transmission lines. The specific acquisition method is as follows: According to the historical information under different climate information, the coefficient of difference in unevenness, the coefficient of difference in average ice thickness, the coefficient of difference in wind direction and the coefficient of difference in wind speed between the representative transmission line and the comparison transmission line are calculated; The weighted unevenness difference coefficient, the average ice thickness difference coefficient, the wind direction difference coefficient and the wind speed difference coefficient are used as the linkage ice dancing probability scores for the representative transmission line and the rest of the transmission lines.

7. The method for predicting ice coating and ice dancing on a power transmission line according to claim 6, characterized in that: The coefficient of unevenness is calculated as follows: in Where, d u is the coefficient of nonuniformity, u rep represents the uneven ice coverage of the transmission line, u rem For the uneven ice coverage of other power lines, represents the length of the transmission line where the ice thickness exceeds the preset value. represents the length of the transmission line where the ice thickness is less than the preset value. For the remaining transmission lines where ice thickness exceeds the preset value, The length of the remaining transmission lines where the ice thickness is less than the preset value; The coefficient of variation of average ice thickness is calculated as follows: in Where, d tk is the average ice thickness variation coefficient, tk erp represents the unit ice mass of the transmission line, tk rem is the unit ice mass of other power lines, m rep represents the quality of the transmission line, L rep represents the length of the transmission line, m rem is the quality of the remaining transmission lines, L rem is the remaining transmission line length.

8. The method for predicting ice coating and ice dancing on a power transmission line according to claim 6, characterized in that: Wind direction difference coefficient, calculated as follows: Where, d θ is the wind direction difference coefficient, n is the historical information number, is the wind direction angle representing the transmission line in the nth history, is the wind direction angle of the remaining transmission lines in the nth history; Wind speed difference coefficient, calculated as follows: Where, d v is the wind speed difference coefficient, n is the historical information number, is the wind speed of the transmission line in the nth history, is the wind speed of the remaining transmission lines in the nth history.

9. The method for predicting ice coating and ice dancing on a power transmission line according to claim 6, characterized in that: Linked ice dancing probability score, the calculation formula is as follows: s=ω1.d u +ω2.d tk +ω3.d θ +ω4.d v Where ω1, ω2, ω3 and ω4 are weights, which are dynamically adjusted in real time according to the diagnosis accuracy of the remaining transmission lines. According to the diagnosis of ice dancing prediction model, the probability of ice dancing on the transmission line is p rep ; Then the probability of ice dancing on the remaining transmission lines is p rem =(1-s)×p rep .

10. The method for predicting ice coating and ice dancing on a power transmission line according to claim 9, characterized in that: The specific methods of real-time dynamic adjustment are as follows: Calculate the prediction results of the remaining transmission lines; If the probability of linked ice dancing between the remaining transmission lines and the representative transmission line is less than a preset value, the difference in unevenness, average ice thickness, wind direction and wind speed between the representative transmission line and the remaining transmission lines is analyzed; Select the attribute with the highest difference ranking as the attribute to be adjusted; Based on the weight corresponding to the attribute to be adjusted, the weight corresponding to the attribute to be adjusted is reduced according to the coefficient ranked first in the preset proportional order; Recalculate the newly generated prediction results of the transmission line. If the probability of ice dancing between the transmission line and the representative transmission line is greater than the preset value, stop adjusting the weight. If the probability of ice dancing between the transmission line and the representative transmission line is greater than the preset value, and the attribute with the highest difference ranking has not changed, the weight corresponding to the attribute to be adjusted is reduced according to the coefficient of the next one in the preset proportional order; If the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value, and the attribute with the highest difference ranking changes, the weight corresponding to the attribute to be adjusted is changed, and the weight corresponding to the attribute to be adjusted is reduced according to the coefficient ranked first in the preset proportional order until the probability of ice dancing in the combined operation between the transmission line and the representative transmission line is greater than the preset value.

Images