CN117074849A - Power system fault emergency processing method based on big data - Google Patents

Abstract

The invention discloses a power system fault emergency processing method based on big data, and relates to the technical field of power system fault processing; the method comprises the following steps: data acquisition, namely acquiring real-time data of each node of the power system; preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data; analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system; and diagnosing faults of the power system based on analysis results, determining fault types and positions, performing emergency treatment, and aiming at the fault types and the fault positions. According to the invention, the circuit nodes in the circuit are subjected to data monitoring, then the standard value and the standard value difference are used for judging, and then the Kalman filtering algorithm is combined for position estimation, so that the monitoring of an omnibearing and full-state power system is realized, and the accuracy of the monitoring and the response speed of fault emergency treatment are increased.

Description

Power system fault emergency processing method based on big data

Technical Field

The invention relates to the technical field of power system fault processing, in particular to a power system fault emergency processing method based on big data.

Background

During the period of using the power system, the power system frequently has fault phenomena of different degrees due to the fact that the borne power load is large and the transmission distance line is long, and after the power system has faults, in order to ensure that a user can use electricity in time, the fault points need to be inspected and overhauled in time.

Through retrieval, the patent with the Chinese patent publication number of CN 110155330A discloses an autonomous line inspection method based on an unmanned aerial vehicle power system, which takes the unmanned aerial vehicle as a carrying main body and comprises a line inspection method of the unmanned aerial vehicle to the power system under strong light and a line inspection method of the unmanned aerial vehicle to the power system under dim light or no light, wherein the line inspection method of the unmanned aerial vehicle to the power system under strong light comprises the following steps: A. line patrol preparation: marking corresponding coordinates of each power tower in an electronic map of a target line patrol area by utilizing the coordinate navigation function of the GPS satellite and a GPS signal positioning module arranged on the power tower; B. line inspection planning: and (C) after forming coordinate points in the electronic map, sequentially connecting the coordinate points in series to form line segments in the electronic map, wherein the line segments are used as flight routes of the line patrol task of the unmanned aerial vehicle.

The above patent suffers from the following disadvantages: the unmanned aerial vehicle inspection system adopts an unmanned aerial vehicle inspection mode to inspect, so that the unmanned aerial vehicle needs to navigate along a line, and only the unmanned aerial vehicle can be inspected at the position, so that fault monitoring is incomplete, and subsequent fault emergency is not timely.

Therefore, the invention provides a power system fault emergency processing method based on big data

Disclosure of Invention

The invention aims to solve the defects in the prior art, and provides a power system fault emergency processing method based on big data.

In order to achieve the above purpose, the present invention adopts the following technical scheme:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

Preferably: in the step S1, the collected node data includes current, voltage and power.

Preferably: in the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

Preferably: in the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

Preferably: in the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

Preferably: in S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

Preferably: in the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +aδ。

Preferably: in the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +aδ。

Preferably: in the step S343, a is gradient, and a is E (0.1, 0.5).

Preferably: in the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state Xv (K-1) =f (K-1) X (K-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

The beneficial effects of the invention are as follows:

1. according to the invention, the circuit nodes in the circuit are subjected to data monitoring, then the standard value and the standard value difference are used for judging, and then the Kalman filtering algorithm is combined for position estimation, so that the monitoring of an omnibearing and full-state power system is realized, and the accuracy of the monitoring and the response speed of fault emergency treatment are increased.

2. According to the invention, through setting the standard value for correction, the standard value can be corrected step by step after the value of the node is judged to be wrong each time according to the normal range data and the standard value data in the fault node, so that the accuracy is further improved.

Drawings

Fig. 1 is a schematic flow chart of a power system fault emergency processing method based on big data.

Detailed Description

The technical scheme of the patent is further described in detail below with reference to the specific embodiments.

Example 1:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.1δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.1δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

Example 2:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed。

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.2δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.2δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

Example 3:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34 of the above-mentioned process,if at a certain point H _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.3δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.3δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state Xv (K-1) =f (K-1) X (K-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

Example 4:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.4δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.4δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

Example 5:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.5δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.5δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

Example 6:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.25δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.25δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.

Example 7:

a power system fault emergency processing method based on big data comprises the following steps:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

In the step S1, the collected node data includes current, voltage and power.

In the step S2, the data preprocessing comprises data cleaning, data drying and data normalization.

In the step S3, the method for analyzing data includes:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Respectively current, voltage and powerIs a standard value of (2);

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

In the step S34, if H at a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

In S34, the step of empirically correcting is as follows:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

In the step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +0.35δ；

In the step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +0.35δ。

In the step S4, the fault location estimation is performed by adopting a kalman filtering algorithm, and specifically includes the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise

According to the invention, the circuit nodes in the circuit are subjected to data monitoring, then the judgment is carried out by utilizing the standard value and the standard value difference, and then the position estimation is carried out by combining the Kalman filtering algorithm, so that the monitoring of an omnibearing and full-state power system is realized, and the accuracy of the monitoring and the response speed of fault emergency treatment are increased.

In addition, the invention can gradually correct the standard value after judging the error of the value of the node each time according to the normal range data and the standard value data in the fault node by setting the standard value correction, thereby further increasing the accuracy.

Claims (10)

1. The power system fault emergency processing method based on big data is characterized by comprising the following steps of:

s1: data acquisition, namely acquiring real-time data of each node of the power system;

s2: preprocessing data, namely preprocessing the acquired data, and ensuring the accuracy and usability of the data;

s3: analyzing the data, namely analyzing the preprocessed data by utilizing a big data analysis technology, and extracting state information and fault characteristics of the power system;

s4: fault diagnosis, which is to diagnose the fault of the power system based on the analysis result and determine the fault type and position;

s5: emergency treatment, namely dispatching maintenance personnel aiming at the fault type and the fault position to carry out fault maintenance.

2. The method for emergency treatment of power system failure based on big data according to claim 1, wherein in step S1, the collected node data includes current, voltage and power.

3. The method for emergency treatment of power system failure based on big data according to claim 1, wherein in step S2, the data preprocessing includes data cleaning, data de-drying and data normalization.

4. The method for emergency treatment of power system failure based on big data according to claim 1, wherein in the step S3, the method for data analysis is as follows:

s31: acquiring data types I, U, P of the acquired data and data corresponding to the data types;

s32: empirically, the manager presets the error range delta _I 、Δ _U 、Δ _P WhereinWherein S is _I 、S _U 、S _P Standard values of current, voltage and power respectively;

s33: data H to be acquired _I 、H _U 、H _P Respectively with S _I 、S _U 、S _P Judgment of H _I 、H _U 、H _P Whether or not to fall into an interval

S34: if the data fall into the interval, the data are judged to be normal, the data and the acquisition time are saved, if the data do not fall into the interval, the data are judged to be abnormal, and the abnormal data, the abnormal data type and the acquisition time are fed back.

5. A base according to claim 4The power system fault emergency processing method of big data is characterized in that in the step S34, if H is a certain point _I 、H _U 、H _P In the process, if abnormal data and normal data exist, then S is performed _I 、S _U 、S _P Experience correction is performed.

6. The method for emergency treatment of power system failure based on big data according to claim 5, wherein in S34, the step of empirically correcting is:

s341: h to be obtained _I 、H _U 、H _P Extracting normal data H from the data _I/U/P ；

S342: setting an empirical formula index value delta, whereinS _I/U/P Is a standard value;

s343: according to H _I/U/P And S is _I/U/P And replacing the standard value according to an empirical formula.

7. The method for emergency treatment of power system failure based on big data as claimed in claim 6, wherein in step S343, if H _I/U/P Greater than S _I/U/P The empirical formula is s=s _I/U/P +aδ。

8. The method for emergency treatment of power system failure based on big data as claimed in claim 7, wherein in step S343, if H _I/U/P Less than S _I/U/P The empirical formula is s=s _I/U/P +aδ。

9. The method for emergency treatment of power system failure based on big data as claimed in claim 8, wherein in step S343, a is gradient, a e (0.1, 0.5).

10. The method for emergency treatment of power system fault based on big data according to claim 1, wherein in step S4, the fault location estimation is performed by using a kalman filter algorithm, and specifically comprises the following steps:

s41: establishing a system state model X (K) =F (K-1) X (K-1) +W (K-1), wherein F (K-1) is a state transition matrix, and W (K-1) is the process noise of the system;

s42: establishing a system measurement value model Y (K) =L (K) X (K) +V (K) wherein L (K) is a measurement matrix and the measurement noise of a V (K) system;

s43: then for time K, the recursive formula of the kalman filter algorithm is:

s431: predicted state X ∈ (k|k-1) =f (K-1) X (K-1|k-1);

s432: prediction error covariance P (k|k-1) =f (K-1) P (K-1|K-1) F (K-1) t+q (K-1), where P (K-1|K-1) is the error covariance at the previous time and Q (K-1) is the covariance of process noise.