Rural power distribution network low voltage cause factor analysis method based on machine learning

Abstract

The invention discloses a rural power distribution network low voltage cause factor analysis method based on machine learning. The analysis method comprises the following steps: acquiring characteristic data of a line in a power distribution area; acquiring power consumption conditions of users of the power distribution network, and constructing a topological structure diagram of the power distribution network; screening out abnormal users in each loop in the power distribution network based on the feature data and the constructed topological structure diagram of the power distribution network; the method comprises the steps of obtaining feature data of an abnormal user and a pre-constructed mathematical model, analyzing voltage loss values corresponding to different influence factors causing low voltage of a power distribution area, and identifying the low voltage leading cause of the power distribution area according to quantitative analysis of the correlation degree between the voltage loss values and a preset voltage evaluation index. Through the analysis method, the leading cause of the low-voltage working condition can be accurately and quickly identified.

Description

Technical Field

[0001] The present invention belongs to the technical field of power distribution networks, and in particular relates to a method for analyzing causes of low voltage in rural power distribution networks based on machine learning. Background Art

[0002] "Low voltage" in the distribution network is a livelihood issue that affects the quality of life of thousands of households. Eliminating "low voltage" is a basic requirement for power supply companies to fulfill their social responsibilities and practice their service purpose, and is an important sign of power supply companies moving towards lean management. The voltage quality of the distribution network is directly related to the safe and economic operation of the power system and the service life of electrical equipment. Too low voltage may cause voltage collapse, causing large-scale power outages, and also reduce the operating capacity of equipment, increase equipment operating energy consumption, burn users' motors, and cause the power of electric lights to drop. Low voltage operation has a great negative impact on both power supply departments and electricity users. Therefore, the voltage quality and safety of the distribution network are difficult problems that power supply companies urgently need to solve.

[0003] In addition to affecting the normal electricity consumption of customers, low voltage problems will also reduce the efficiency and economic benefits of electrical appliances, affect the normal operation of production equipment and product quality, increase power loss and energy loss of the power grid, and endanger the safe operation of the power system and power supply and use equipment. By scientifically and comprehensively analyzing the causes of low voltage problems in rural areas and taking targeted governance measures, we can avoid the waste of manpower, money and materials caused by blind investment and repeated transformation in the process of low voltage governance, and effectively solve the existing low voltage problems.

[0004] At present, the causes of low voltage in rural distribution networks are mainly analyzed through manual experience. However, judging the causes of low voltage by manual experience is inefficient and cannot accurately capture the main causes of low voltage. Summary of the invention

[0005] In view of this, the purpose of the present invention is to provide a method for analyzing the causes of low voltage in rural distribution networks based on machine learning, aiming to solve the technical problems of low efficiency and low accuracy of current low voltage cause analysis methods.

[0006] To achieve the above object, the present invention provides the following technical solution: a method for analyzing the causes of low voltage in rural distribution networks based on machine learning, comprising the following steps:

[0007] Obtain characteristic data of lines in the distribution area;

[0008] Obtain the power consumption of distribution network users and construct a distribution network topology diagram;

[0009] Screening out abnormal users in each loop of the distribution network based on the characteristic data and the constructed distribution network topology diagram;

[0010] Acquire the characteristic data of abnormal users and the pre-built mathematical model, analyze the voltage loss values ​​corresponding to the different influencing factors that cause low voltage in the distribution station area, and identify the dominant cause of low voltage in the distribution station area based on the quantitative analysis of the correlation between the voltage loss value and the preset voltage evaluation index.

[0011] Furthermore, the characteristic data in the step of obtaining characteristic data of the lines in the distribution station area includes voltage and current values ​​at the low-voltage outlet end of the distribution station area, and voltage, current and power factor values ​​of all users in the distribution station area.

[0012] Furthermore, the step of obtaining the power consumption of the distribution network users and constructing the distribution network topology diagram includes:

[0013] Conduct on-site survey of the distribution area line map to identify the distribution of all distribution users;

[0014] Obtain the high-power electricity consumption of distribution users, and based on the screened high-power distribution users, study their on-site geographical environment and the phase of the user voltage, and construct a distribution network topology diagram.

[0015] Furthermore, the high power means that the user power is not less than 2000W.

[0016] Furthermore, the abnormal user satisfies at least one of the following conditions:

[0017] The voltage in each voltage loop is the lowest; the power supply radius in each voltage loop is the longest; and the wire diameter in each voltage loop is the smallest.

[0018] Furthermore, the characteristic data of the abnormal user include the user's name; the user's phase; the length and wire diameter of the main line from the low-voltage outlet of the transformer to the user; the length and wire diameter of the branch line to the user; the length and wire diameter of the user's downline line; the user's power factor value; the user's lowest voltage value, the user's current value in the same period, the current value of the phase with the largest load at the low-voltage outlet end of the substation, the current value of the phase with the smallest load at the low-voltage outlet end of the substation, the phase voltage of the low-voltage outlet end of the phase where the user is located; and the low voltage collection time.

[0019] Furthermore, the pre-built mathematical model is determined as follows:

[0020]

[0021] Where ΔU is the voltage loss value on the low-voltage side of the distribution transformer; S e is the rated capacity of the distribution transformer; ε is the three-phase unbalance; S is the conductor cross-sectional area, and L is the conductor length; is the power factor angle; β is the load rate; U N is the terminal voltage.

[0022] Furthermore, the method further comprises the following steps:

[0023] Sort out low voltage on-site problems and analyze their causes, and determine the weights corresponding to each voltage evaluation indicator;

[0024] According to the weights corresponding to the various voltage evaluation indicators and the voltage loss value, further optimizing and checking the correlation between the various voltage evaluation indicators and the voltage loss value in the distribution station area;

[0025] The correlation degree is used to quantitatively sort the voltage evaluation indicators according to their influence, so as to determine the main causes of the low voltage phenomenon corresponding to the distribution station area.

[0026] Furthermore, based on identifying the dominant cause of low voltage in the distribution substation area, control measures for the distribution substation area are determined, including:

[0027] (1) When the main cause of the low voltage is the unbalanced three-phase current, at least one of the following treatment measures is adopted: manual phase adjustment, setting of phase adjustment switch, three-phase four-wire extension, and setting of three-phase unbalance adjustment device;

[0028] (2) When the main cause of the low voltage is the small cross-sectional area of ​​the wire, the wire should be replaced by a wire with a smaller cross-sectional area.

[0029] The beneficial effect of the present invention is that compared with the prior art, the method for analyzing the causes of low voltage in rural distribution networks in the present invention obtains characteristic data of lines in distribution network areas; obtains the power consumption of distribution network users, and constructs a distribution network topology diagram; screens out abnormal users in each circuit of the distribution network based on the characteristic data and the constructed distribution network topology diagram; obtains the characteristic data of abnormal users and a pre-constructed mathematical model, analyzes the voltage loss values ​​corresponding to different influencing factors causing low voltage in distribution network areas, and identifies the dominant causes of low voltage in distribution network areas based on a quantitative analysis of the correlation between the voltage loss values ​​and the preset voltage evaluation index. Through this analysis method, the dominant causes of low voltage conditions can be accurately and quickly identified.

[0030] Other advantages, objectives and features of the present invention will be described in the following description and will be apparent to those skilled in the art to some extent, or those skilled in the art may be taught from the practice of the present invention. The objectives and other advantages of the present invention may be realized and obtained through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] In order to make the purpose, technical solution and beneficial effects of the present invention clearer, the present invention provides the following drawings for illustration:

[0032] Figure 1 A flowchart of a method for analyzing causes of low voltage in rural distribution networks based on machine learning proposed in one embodiment of the present invention;

[0033] Figure 2 A topological diagram constructed in a method for analyzing causes of low voltage in a rural distribution network is proposed in one embodiment of the present invention;

[0034] Figure 3 This is the loss ranking of the causal factors after simulation calculation in the simulation software proposed in one embodiment of the present invention. DETAILED DESCRIPTION

[0035] like Figure 1 to Figure 2 As shown, this embodiment proposes a method for analyzing the causes of low voltage in rural distribution networks based on machine learning, comprising the following steps:

[0036] S01: Obtain characteristic data of the lines in the distribution area;

[0037] S02: Obtain the power consumption of users in the distribution network and construct a topological structure diagram of the distribution network;

[0038] S03: Screening out abnormal users in each loop of the distribution network based on the characteristic data and the constructed distribution network topology diagram;

[0039] S04: Acquire characteristic data of abnormal users and a pre-built mathematical model, analyze the voltage loss values ​​corresponding to different influencing factors causing low voltage in the distribution station area, and identify the dominant cause of low voltage in the distribution station area based on quantitative analysis of the correlation between the voltage loss value and the preset voltage evaluation index.

[0040] In the present application, by acquiring the characteristic data of the lines in the distribution station area and constructing a power grid topology diagram, abnormal users in each circuit in the distribution network can be screened out, normal users can be eliminated, and the characteristic data of abnormal users can be analyzed in a targeted manner, thereby greatly improving the analysis efficiency; by using a low-voltage mathematical model to tune and train the machine learning model, a data model is obtained, and the characteristic data of the acquired abnormal users is input into the data model for calculation, and the voltage loss values ​​corresponding to the different influencing factors of the low voltage in the distribution station area are analyzed, and the obtained voltage loss values ​​are quantitatively analyzed in correlation with the voltage evaluation indicators preset in the simulation software, thereby accurately and quickly identifying the dominant causes of low voltage conditions.

[0041] Furthermore, the characteristic data described in the step of obtaining characteristic data of the line in the distribution substation includes the voltage and current values ​​at the low-voltage outlet of the distribution substation, and the voltage, current, and power factor values ​​of all users in the distribution substation. Listed in this application are the main factors affecting low voltage. Of course, according to actual conditions and specific needs, the characteristic data also includes other data values, which are not limited here.

[0042] In this application, the causes of low voltage are roughly as follows: 1. Large power supply radius; 2. Small conductor cross-sectional area; 3. Unbalanced three-phase current; 4. High transformer load rate; 5. Low user power factor; 6. Low tap switch gear or low voltage on the 10 kV high-voltage side; 7. Line aging, tree barriers; 8. Holidays or seasonal heavy loads, etc. On the basis of the above limited reasons, by obtaining the above characteristic data of the lines in the distribution station area, find out the peak power consumption period, and simulate and analyze the transformer capacity of the station area, the length and wire diameter of the main line and branch line, the length and wire diameter of the abnormal user's down-line, the lowest voltage value, the user's current value, power factor, the voltage and current value at the low-voltage outlet of the station area at the corresponding time, so as to accurately and quickly identify the dominant cause of the low voltage condition.

[0043] Furthermore, the step of obtaining the power consumption of the distribution network users and constructing the distribution network topology diagram includes:

[0044] Conduct on-site survey of the distribution area line map to identify the distribution of all distribution users;

[0045] Obtain the high-power electricity consumption of distribution users, and based on the screened high-power distribution users, study their on-site geographical environment and the phase of the user voltage, and construct a distribution network topology diagram.

[0046] In this application, through on-site surveys, the distribution of all distribution users is summarized, and the users' high-power electricity consumption is screened out from all distribution users, so as to understand the on-site geographical environment of the high-power distribution users and the phase of the user's voltage, that is, the user's voltage is in phase A, phase B, and phase C, so as to construct a distribution network topology diagram.

[0047] In actual use, by analyzing the user's high-power electricity consumption, such as whether the user has high-power air conditioners or high-power mechanical equipment at home and the number of such equipment, we can understand the phase type of the user, determine the power situation of the phase type, and construct a distribution network topology diagram, which is conducive to the subsequent adjustment of the power distribution situation under each phase.

[0048] Preferably, the high power is a user power of not less than 2000W. When less than 2000W is the normal power used by the user, when greater than or equal to 2000W, this is used as the base number as the data source for constructing the distribution network topology diagram. Of course, in this application, according to actual conditions and specific needs, the value of the user's high power can also be adjusted accordingly, such as 2200W, 3300W, 3500W, etc., which is not the only limitation here.

[0049] Preferably, the abnormal user satisfies at least one of the following conditions:

[0050] The voltage in each voltage loop is the lowest; the power supply radius in each voltage loop is the longest; and the wire diameter in each voltage loop is the smallest.

[0051] Specifically, if the user has the lowest voltage in the voltage loop, or the longest power supply radius in the voltage loop, or the smallest wire diameter in the voltage loop, or the lowest voltage in the voltage loop and the longest power supply radius in the voltage loop, or the lowest voltage in the voltage loop and the smallest wire diameter in the voltage loop, or the longest power supply radius in the voltage loop and the smallest wire diameter in the voltage loop, or the lowest voltage in the voltage loop, the longest power supply radius in the voltage loop, and the smallest wire diameter in the voltage loop, then the user is marked as an abnormal user.

[0052] Preferably, the characteristic data of the abnormal user include the user's name; the user's phase; the length and wire diameter of the main line from the low-voltage outlet of the transformer to the user; the length and wire diameter of the branch line to the user; the length and wire diameter of the user's downline line; the user's power factor value; the user's lowest voltage value, the user's current value in the same period, the current value of the phase with the largest load at the low-voltage outlet end of the substation, the current value of the phase with the smallest load at the low-voltage outlet end of the substation, the phase voltage of the low-voltage outlet end of the phase where the user is located; and the low voltage collection time.

[0053] By taking the characteristic data of abnormal users as the research object, the low-voltage mathematical model is used to tune and train the machine learning model to obtain the data model. By inputting the above-mentioned characteristic data of abnormal users into the data model for calculation, the voltage loss values ​​corresponding to the different influencing factors of low voltage in the distribution station area are analyzed and obtained.

[0054] Preferably, the pre-built mathematical model is determined as follows:

[0055]

[0056] Where ΔU is the voltage loss value on the low-voltage side of the distribution transformer; S e is the rated capacity of the distribution transformer; ε is the three-phase unbalance; S is the conductor cross-sectional area, and L is the conductor length; is the power factor angle; β is the load rate; U Nis the terminal voltage.

[0057] In the present application, the voltage loss value can be calculated by measuring the values ​​of the above variables.

[0058] Preferably, the voltage evaluation index preset in the simulation software includes the length of the power supply wire, the size of the wire cross-section, the load rate of the distribution transformer, the three-phase current imbalance, the power factor, the terminal voltage, and the rated capacity of the distribution transformer. That is, the preset voltage evaluation index is consistent with the independent variable of the constructed mathematical model. In this way, by adjusting the value of the evaluation index, the influence of each evaluation index on the voltage loss value can be analyzed, and the quantitative relationship between the two can be sought, thereby confirming the dominant cause of low voltage in the distribution station area.

[0059] Furthermore, the acquisition of characteristic data of abnormal users and a pre-built mathematical model, analyzing the voltage loss values ​​corresponding to different influencing factors causing low voltage in the distribution station area, and identifying the dominant cause of low voltage in the distribution station area based on the quantitative analysis of the correlation between the voltage loss value and the preset voltage evaluation index also includes the following steps:

[0060] Sort out low voltage on-site problems and analyze their causes, and determine the weights corresponding to each voltage evaluation indicator;

[0061] According to the weights corresponding to the various voltage evaluation indicators and the voltage loss value, further optimizing and checking the correlation between the various voltage evaluation indicators and the voltage loss value in the distribution station area;

[0062] The correlation degree is used to quantitatively sort the voltage evaluation indicators according to their influence, so as to determine the main causes of the low voltage phenomenon corresponding to the distribution station area.

[0063] In this application, by sorting out low-voltage on-site problems and analyzing the causes of the problems, the voltage evaluation indicators are corrected and the weights corresponding to each voltage evaluation indicator are determined. According to the corrected voltage evaluation indicators and voltage loss values, the correlation between each voltage evaluation indicator and the voltage loss value can be further optimized. The correlation is used to quantitatively rank the above-mentioned voltage evaluation indicators and determine the main causes of the low voltage phenomenon corresponding to the distribution station area.

[0064] In this application, by improving and correcting the mathematical model, the quantitative relationship and mathematical model between voltage deviation and indicators such as power supply radius, conductor cross-section size, distribution transformer load rate, three-phase current imbalance, power factor, etc. are found, so as to determine the main cause of low voltage.

[0065] In this application, based on identifying the dominant cause of low voltage in the distribution substation area, the control measures for the distribution substation area are determined, which specifically include:

[0066] (1) When the main cause of the low voltage is the unbalanced three-phase current, at least one of the following treatment measures is adopted: manual phase adjustment, setting of phase adjustment switch, three-phase four-wire extension, and setting of three-phase unbalance adjustment device;

[0067] (2) When the main cause of the low voltage is the small cross-sectional area of ​​the wire, the wire should be replaced by a wire with a smaller cross-sectional area.

[0068] In this application, the evaluation indicators affecting low voltage can be prioritized under the multi-factor analysis dimension through simulation software, such as Figure 3 As shown, the primary factor causing low voltage is the unbalanced three-phase current, followed by the small cross-sectional area of ​​the wire. After confirming the main influencing factors, the corresponding countermeasures can be matched, and then the solution to the low voltage can be determined, which is accurate and efficient.

[0069] Finally, it should be noted that the above preferred 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 through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. A method for analyzing the causes of low voltage in rural distribution networks based on machine learning, characterized in that: The steps include: Obtain characteristic data of lines in the distribution area; Obtain the power consumption of distribution network users and construct a distribution network topology diagram; Screening out abnormal users in each loop of the distribution network based on the characteristic data and the constructed distribution network topology diagram; Acquire the characteristic data of abnormal users and the pre-built mathematical model, analyze the voltage loss values ​​corresponding to the different influencing factors that cause low voltage in the distribution station area, and identify the dominant cause of low voltage in the distribution station area based on the quantitative analysis of the correlation between the voltage loss value and the preset voltage evaluation index.

2. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1 is characterized in that: The characteristic data in the step of obtaining characteristic data of the lines in the distribution station area include the voltage and current values ​​at the low-voltage outlet end of the distribution station area, and the voltage, current and power factor values ​​of all users in the distribution station area.

3. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1 is characterized in that: The steps of obtaining the power consumption of the users of the distribution network and constructing the distribution network topology diagram include: Conduct on-site survey of the distribution area line map to identify the distribution of all distribution users; Obtain the high-power electricity consumption of distribution users, and based on the screened high-power distribution users, study their on-site geographical environment and the phase of the user voltage, and construct a distribution network topology diagram.

4. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 3 is characterized in that: The high power mentioned above means that the user power is not less than 2000W.

5. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1 is characterized in that: The abnormal user meets at least one of the following conditions: The voltage in each voltage loop is the lowest; the power supply radius in each voltage loop is the longest; and the wire diameter in each voltage loop is the smallest.

6. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1 is characterized in that: The characteristic data of the abnormal user include the user's name; the user's phase; the length and wire diameter of the main line from the low-voltage outlet of the transformer to the user; the length and wire diameter of the branch line to the user; the length and wire diameter of the user's downline line; the user's power factor value; the user's lowest voltage value, the user's current value in the same period, the current value of the phase with the largest load at the low-voltage outlet end of the substation, the current value of the phase with the smallest load at the low-voltage outlet end of the substation, the phase voltage of the low-voltage outlet end of the phase where the user is located; and the low voltage collection time.

7. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1 is characterized in that: The pre-built mathematical model is determined as follows: Where ΔU is the voltage loss value on the low-voltage side of the distribution transformer; S e is the rated capacity of the distribution transformer; ε is the three-phase unbalance; S is the conductor cross-sectional area, and L is the conductor length; is the power factor angle; β is the load rate; U N is the terminal voltage.

8. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1 is characterized in that: The following steps are also included: Sort out low voltage on-site problems and analyze their causes, and determine the weights corresponding to each voltage evaluation indicator; According to the weights corresponding to the various voltage evaluation indicators and the voltage loss value, further optimizing and checking the correlation between the various voltage evaluation indicators and the voltage loss value in the distribution station area; The correlation degree is used to quantitatively sort the voltage evaluation indicators according to their influence, so as to determine the main causes of the low voltage phenomenon corresponding to the distribution station area.

9. The method for analyzing the causes of low voltage in rural distribution networks based on machine learning according to claim 1, characterized in that: Based on the identification of the main causes of low voltage in the distribution substation area, the control measures for the distribution substation area are determined, including: (1) When the main cause of the low voltage is the unbalanced three-phase current, at least one of the following treatment measures is adopted: manual phase adjustment, setting of phase adjustment switch, three-phase four-wire extension, and setting of three-phase unbalance adjustment device; (2) When the main cause of the low voltage is the small cross-sectional area of ​​the wire, the wire should be replaced by a wire with a smaller cross-sectional area.

Images