CN104037943A - Method and system for monitoring voltage and capable of improving power grid voltage quality - Google Patents

Abstract

The invention provides a method for monitoring voltage and capable of improving power grid voltage quality. The method for monitoring the voltage and capable of improving the power grid voltage quality includes: obtaining real time voltage data and voltage statistic data of a plurality of monitoring points; analyzing history status of voltage quality indexes according to the collected voltage statistic data of all the monitoring points; performing online prediction on future trends of the voltage quality indexes according to the current collected real time voltage data of all the monitoring points. A system for monitoring the voltage and capable of improving the power grid voltage quality, which is used to achieve the method, comprises a data acquisition unit, a history time period voltage quality index analysis unit, a future time period voltage quality index prediction unit and a data storage unit. The method and the system for monitoring the voltage and capable of improving the power grid voltage quality achieve deep analysis for the voltage quality and timely judgment for the future trends of the power grid voltage quality, and improve the effective degree of power grid voltage quality monitoring.

Description

A voltage monitoring method and system for improving grid voltage quality

technical field

The invention relates to the technical field of grid voltage quality monitoring, in particular to a voltage monitoring method and system for improving grid voltage quality.

Background technique

With the sustained and rapid development of the national economy and the continuous improvement of people's living standards, the trend of rapid growth in my country's electricity demand will not change for a long time. With the development of the power grid and the continuous increase of power supply load, if the voltage fluctuation exceeds the allowable range and lasts for a long time, it will cause great harm: for the power supply grid, low voltage will affect the ability of power generation and power supply equipment, and affect the reliability of power supply; The voltage of the equipment is unstable, which will affect the service life, or even burn out, increasing the line loss. The above-mentioned reasons will also have adverse effects on the normal electricity consumption of power users. Therefore, it is particularly important to strengthen real-time monitoring of grid voltage and data management and analysis to improve voltage quality.

At present, there are some patents on the monitoring and analysis of voltage quality, such as "200910164280.9 (an on-line monitoring method for grid voltage stability based on local indicators of voltage stability)" according to the grid topology parameters and related The node voltage and current phasor calculates the equivalent system parameters of single power supply power transmission at the monitored point, judges and warns the voltage stability of the monitored point, and effectively realizes the online real-time monitoring of the grid voltage stability. "201210034024.X (a voltage quality monitoring method)" first collects the voltage quality data of each voltage level of the regional power grid, through the analysis of the data, adopts different measures to analyze and judge the power supply voltage quality indicators, and gives corresponding adjustments It is beneficial to deal with adaptive analysis occasions. "201310376476.0 (a wide-range self-adaptive voltage quality monitoring method)" calculates the voltage deviation rate between the adopted voltage and each nominal voltage, intelligently analyzes the continuously calculated voltage deviation rate results, and deduces the rated voltage of the current system. Realize self-adaptation of voltage monitoring range, and complete voltage quality monitoring without manual intervention. "201310528971.9 (A method for online prediction of static voltage stability limit of power system)" based on Lasso's sample dimensionality reduction method, self-organizing feature map network sample screening method and error backpropagation neural network to predict the static voltage stability limit of power system Offline training and online prediction can effectively improve the offline training efficiency and online prediction effect of the error backpropagation neural network. The above-mentioned patents do not conduct an in-depth analysis of the grid voltage quality index, nor make a timely judgment on the future trend of the grid voltage quality, so it is difficult to effectively monitor the grid voltage quality.

Contents of the invention

Aiming at the deficiencies in the prior art, the invention provides a voltage monitoring method and system for improving the voltage quality of a grid.

Technical scheme of the present invention is:

A voltage monitoring method for improving grid voltage quality, comprising the following steps:

Step 1: Obtain real-time voltage data and voltage statistical data of multiple monitoring points;

The voltage statistical data includes daily voltage statistical data, monthly voltage statistical data, quarterly voltage statistical data, annual voltage statistical indicators and indicator statistical data of departments at all levels at monitoring points;

Among them, the monthly statistical data of voltage includes the statistical value of voltage in the current month, the statistical value of voltage in the previous month, the typical day and time data of the current month, the typical day and time data of the previous month, the voltage reliability data of the current month, the voltage reliability data of the previous month, the statistical value of power outages in the current month and the previous month. Statistical value of monthly outages; voltage annual statistical indicators include voltage annual pass rate, voltage annual over-limit rate and voltage annual over-low limit rate; indicator statistics of departments at all levels include daily pass rate statistics, monthly pass rate statistics, Quarterly pass rate statistics and annual pass rate statistics;

Step 2: According to the collected voltage statistics data of each monitoring point, analyze the historical situation of the voltage quality index;

Step 2-1: Select the monitoring points to be analyzed;

Step 2-2: Select the voltage quality index to be analyzed. The voltage quality index includes voltage qualification rate, voltage standard deviation and voltage probability density;

Step 2-3: Select the analysis method of year-by-quarter, year-by-month, quarter-by-month or month-by-day, and select a specific historical time period according to the selected analysis method;

Step 2-4: Calculate the voltage quality index value of the selected monitoring point in the historical time period according to the selected analysis method and historical time period;

Step 2-5: Display the voltage quality index value of the selected monitoring point in the form of table, curve and bar graph;

Step 2-6: According to the voltage quality index value of the selected monitoring point and the setting value of the voltage quality index of each level, determine the monitoring point with poor voltage quality index, and send an alarm so that the electric power personnel can monitor the abnormal voltage quality of the monitoring point. diagnostic treatment;

Step 3: According to the current collected real-time voltage data of each monitoring point, online prediction of the future trend of the voltage quality index;

Step 3-1: Select the monitoring points that need to be predicted;

Step 3-2: Select the voltage quality index that needs to be predicted;

Step 3-3: Select the forecast method of year by quarter, year by month, quarter by month or month by day, and select a specific future time period according to the selected forecast method;

Step 3-4: When the voltage quality index data of the selected monitoring point is missing in the historical time period, supplement the missing voltage quality index data;

Step 3-5: Establish an online prediction model of the future trend of the voltage quality index. The input of the model is the value of the voltage quality index within the time range from the current moment to a certain historical moment, and the output of the model is the future to be predicted The voltage quality index value of the time period;

Step 3-6: Predict the future trend of the voltage quality index of the selected monitoring point according to the established online prediction model of the future trend of the voltage quality index;

Step 3-7: Display the online prediction results of the future trend of the voltage quality indicators of the selected monitoring points in the form of tables, curves and bar graphs;

Step 3-8: According to the online prediction results of the future trend of the voltage quality indicators of the selected monitoring points and the set values of the voltage quality indicators of each level, determine the monitoring points with poor voltage quality indicators at the future time, and give early warning, so that the electric power personnel can monitor the existing voltage The monitoring points of hidden quality problems are dealt with in advance.

In steps 3-5, the future trend online prediction model of voltage quality index is established according to the following steps:

Step 3-5-1: Establish a time series model. The input of the time series model is the voltage quality index value within the time range from the current moment to a certain historical moment, and the output of the time series model is the future time period to be predicted. Voltage quality index value;

Step 3-5-2: Establish a gray model, the input of the gray model is the voltage quality index value in the time range from the current moment to a certain historical moment, and the output of the gray model is the voltage quality in the future time period to be predicted Index value;

Step 3-5-3: Establish a combined forecasting model, the input of the combined forecasting model includes the voltage quality index value of the future time period predicted by the time series model and the voltage quality index value of the future time period predicted by the gray model, and the output of the combined forecasting model It is the weighted sum of the voltage quality index value of the future time period predicted by the time series model and the voltage quality index value of the future time period predicted by the gray model, that is, the voltage quality index value of the future time period predicted by the combined forecasting model;

Step 3-5-4: Establish a BP neural network model. The input of the BP neural network model is the voltage quality index value of the future time period predicted by the combined forecasting model, and the output of the BP neural network model is the value of the future time period predicted by the combined forecasting model. The error of the voltage quality index value;

Step 3-5-5: Establish the future trend online prediction model of the voltage quality index, the input of the model includes the voltage quality index value of the future time period predicted by the combined prediction model and the future time predicted by the combined prediction model output by the BP neural network model The error of the voltage quality index value of the segment, the output of the online prediction model is the error of the voltage quality index value of the future time period predicted by the combined prediction model and the voltage quality index value of the future time period predicted by the combined prediction model output by the BP neural network model The sum is the prediction result of the voltage quality index.

The method of selecting monitoring points to be analyzed in step 2-1 and the method of selecting monitoring points to be predicted in step 3-1 both include selecting a single monitoring point, selecting multiple monitoring points, and selecting all relevant monitoring points of a line at one time One-time selection of all monitoring points of the substation and one-time selection of all monitoring points of the county power supply bureau.

The voltage quality index includes voltage qualification rate, voltage standard deviation and voltage probability density.

Select a specific historical time period according to the selected analysis method described in steps 2-3. When the analysis method is year by quarter, select the historical time period as year; when the analysis method is year by month, select the historical time period as Year; when the analysis method is quarter by month, the historical time period is selected as year and quarter; when the analysis method is month by day, the historical time period is selected as year and month.

In step 3-3, select a specific future time period based on the selected forecasting method. When the forecasting method is yearly and quarterly, the forecast result belongs to the remaining quarter of the year; when the monitoring method is yearly and monthly, forecast The time of the result is the remaining month of the current year; when the monitoring method is quarter by month, the time of the forecast result is the remaining month of the quarter; when the monitoring method is month by day, the time of the forecast result is the rest of the month number of days.

The cubic spline interpolation method is used to supplement the missing voltage quality index data described in steps 3-4.

A power grid voltage quality monitoring system for implementing the monitoring method for improving power grid voltage quality, including a data acquisition unit, a voltage quality index analysis unit in a historical time period, a voltage quality index prediction unit in a future time period, and a data storage unit;

The data acquisition unit is used to acquire real-time voltage data and voltage statistical data of the monitoring point;

The voltage quality index analysis unit in the historical time period is used to analyze the historical situation of the voltage quality index according to the statistical data of each monitoring point obtained by the data acquisition unit, including: selecting the monitoring point to be analyzed and the voltage quality index to be analyzed; selecting According to the analysis method of year-by-quarter, year-by-month, quarter-by-month or month-by-day and select a specific historical time period according to the selected analysis method; according to the selected analysis method and historical time period, calculate the selected monitoring points at the historical time The voltage quality index value of the segment; the voltage quality index value of the selected monitoring point is displayed in the form of table, curve and bar graph; the voltage quality is determined according to the voltage quality index value of the selected monitoring point and the voltage quality index setting value of each level Alarms will be sent to monitoring points with poor indicators;

The voltage quality index prediction unit in the future time period is used to predict the future trend of the voltage quality index online according to the real-time voltage data of each monitoring point currently obtained, including: selecting the monitoring point to be predicted and the voltage quality index to be predicted; selecting Yearly by quarter, year by month, quarter by month or month by day, and select a specific future time period according to the selected forecast method; when the voltage quality index data of the selected monitoring point in the historical time period is missing, Then supplement the missing voltage quality index data; establish an online prediction model of the future trend of the voltage quality index, the input of the model is the voltage quality index value in the time range from the current moment to a certain historical moment, the model’s The output is the voltage quality index value of the future time period to be predicted; according to the established online prediction model of the future trend of the voltage quality index, the future trend of the voltage quality index of the selected monitoring point is predicted; in the form of tables, curves and bar graphs Display the online prediction result of the future trend of the voltage quality index of the selected monitoring point; according to the online prediction result of the future trend of the voltage quality index of the selected monitoring point and the set value of the voltage quality index of each level, determine the probability of the poor voltage quality index in the future time period Monitoring points for early warning;

The data storage unit is used to store the real-time voltage data and voltage statistical data of the monitoring point, the online prediction model parameters of the future trend of the voltage quality index, the value of each voltage quality index of the monitoring point in each historical time period, and the value of the monitoring point in the future time period. Predicted value of voltage quality index.

The historical time period voltage quality index analysis unit includes a historical time period voltage quality index value calculation module and a historical time period voltage quality index analysis display and alarm module;

The voltage quality index value calculation module in the historical time period is used to select the monitoring point to be analyzed, the voltage quality index and the analysis method to be analyzed, select a specific historical time period according to the selected analysis method, and then calculate the selected monitoring point in the historical time period. The voltage quality index value of the segment;

The voltage quality index analysis display and alarm module in the historical time period is used to display the voltage quality index value of the selected monitoring point in the form of table, curve and bar graph; according to the voltage quality index value of the selected monitoring point and the voltage quality index setting of each level Set the value to determine the monitoring points with poor voltage quality indicators to alarm.

The voltage quality index prediction unit in the future time period includes a monitoring point historical data supplement module, a time series model establishment module, a gray model establishment module, a combined forecast model establishment module, a BP neural network model establishment module, a voltage quality index prediction module, a voltage quality Index prediction result display and early warning module;

The monitoring point historical data supplement module is used to select the monitoring points that need to be predicted, the voltage quality indicators that need to be predicted, and the prediction method, and select a specific future time period according to the selected prediction method. When the index data is missing, the missing voltage quality index data will be supplemented;

The time series model building module is used to establish a time series model that takes the voltage quality index value in the time range from the current moment to a certain historical moment as input, and outputs the voltage quality index value in the future time period to be predicted;

The gray model establishment module is used to establish a gray model that takes the voltage quality index value in the time range from the current moment to a certain historical moment as input, and outputs the voltage quality index value in the future time period to be predicted;

The combined prediction model building module is used to establish the voltage quality index value of the future time period predicted by the time series model and the voltage quality index value of the future time period predicted by the gray model. Input, the voltage quality of the future time period predicted by the time series model The weighted sum of the index value and the voltage quality index value of the future time period predicted by the gray model is the output combined prediction model;

The BP neural network model building module is used to input the voltage quality index value of the future time period predicted by the combined forecasting model, and output the BP neural network model with the error of the voltage quality index value of the future time period predicted by the combined forecasting model;

The voltage quality index prediction module is used to establish the error of the voltage quality index value of the future time period predicted by the combination prediction model and the voltage quality index value of the future time period predicted by the BP neural network model output as input, and the combined prediction The sum of the error sum of the voltage quality index value predicted by the model in the future time period and the combined prediction model output by the BP neural network model is the future trend online prediction model of the output voltage quality index, and according to The established online prediction model of the future trend of the voltage quality index predicts the future trend of the voltage quality index of the selected monitoring point, and obtains the prediction result of the voltage quality index;

The voltage quality index prediction result display and early warning module is used to display the online prediction results of the future trend of the voltage quality index of the selected monitoring point in the form of tables, curves and bar graphs, and according to the online prediction of the future trend of the voltage quality index of the selected monitoring point Prediction results, the set values of voltage quality indicators at each level, determine the monitoring points of poor voltage quality indicators in the future, and carry out early warning.

Beneficial effect:

Aiming at the problem that the current grid voltage quality monitoring method does not conduct in-depth analysis of the voltage quality and does not make timely judgments on the future trend of the grid voltage quality, it is difficult to effectively monitor the grid voltage quality. The present invention proposes a method based on improving the grid voltage Quality monitoring methods, including data acquisition, monitoring of voltage quality indicators at monitoring points at historical moments, and prediction of voltage quality indicators at monitoring points at future moments based on a combined prediction model and BP neural network model. The in-depth analysis of the voltage quality and the timely judgment of the future trend of the grid voltage quality are realized, and the effectiveness of the grid voltage quality monitoring is improved.

Description of drawings

Fig. 1 is the structural block diagram of the monitoring system that improves grid voltage quality of the embodiment of the present invention;

Fig. 2 is the flowchart of the monitoring method that improves grid voltage quality according to the embodiment of the present invention;

Fig. 3 is the month-by-year monitoring result curve of the monitoring point voltage standard deviation of the monitoring method for improving the grid voltage quality according to the specific embodiment of the present invention;

Fig. 4 is the year by month monitoring result bar graph of the monitoring point voltage standard deviation of the monitoring method of improving grid voltage quality of the specific embodiment of the present invention;

Fig. 5 is the yearly and monthly monitoring result curve of two monitoring point voltage pass rates of the monitoring method for improving grid voltage quality according to a specific embodiment of the present invention;

Fig. 6 is the yearly and monthly monitoring results of two monitoring point voltage pass rates of a monitoring method for improving grid voltage quality according to a specific embodiment of the present invention;

Fig. 7 is the flowchart of analyzing the historical situation of the voltage quality index according to the specific embodiment of the present invention;

Fig. 8 is the flow chart that carries out online prediction to the future trend of voltage quality index of the specific embodiment of the present invention;

FIG. 9 is a flow chart of establishing an online prediction model for future trends of voltage quality indicators according to a specific embodiment of the present invention.

Detailed ways

The specific implementation manners of the present invention will be described in detail below in conjunction with the accompanying drawings.

In this embodiment, the grid voltage quality monitoring system of the monitoring method for improving grid voltage quality, as shown in Figure 1, includes a data acquisition unit, a voltage quality index analysis unit in a historical time period, a voltage quality index prediction unit in a future time period, and a data storage unit. unit;

The data acquisition unit is used to acquire real-time voltage data and voltage statistical data of the monitoring point;

The voltage quality index analysis unit in the historical time period is used to analyze the historical situation of the voltage quality index according to the statistical data of each monitoring point obtained by the data acquisition unit, including: selecting the monitoring point to be analyzed and the voltage quality index to be analyzed; selecting According to the analysis method of year-by-quarter, year-by-month, quarter-by-month or month-by-day and select a specific historical time period according to the selected analysis method; according to the selected analysis method and historical time period, calculate the selected monitoring points at the historical time The voltage quality index value of the segment; the voltage quality index value of the selected monitoring point is displayed in the form of table, curve and bar graph; the voltage quality is determined according to the voltage quality index value of the selected monitoring point and the voltage quality index setting value of each level Alarms will be sent to monitoring points with poor indicators;

The voltage quality index prediction unit in the future time period is used to predict the future trend of the voltage quality index online according to the real-time voltage data of each monitoring point currently obtained, including: selecting the monitoring point to be predicted and the voltage quality index to be predicted; selecting Yearly by quarter, year by month, quarter by month or month by day, and select a specific future time period according to the selected forecast method; when the voltage quality index data of the selected monitoring point in the historical time period is missing, Then supplement the missing voltage quality index data; establish an online prediction model of the future trend of the voltage quality index, the input of the model is the voltage quality index value in the time range from the current moment to a certain historical moment, the model’s The output is the voltage quality index value of the future time period to be predicted; according to the established online prediction model of the future trend of the voltage quality index, the future trend of the voltage quality index of the selected monitoring point is predicted; in the form of tables, curves and bar graphs Display the online prediction result of the future trend of the voltage quality index of the selected monitoring point; according to the online prediction result of the future trend of the voltage quality index of the selected monitoring point and the set value of the voltage quality index of each level, determine the probability of the poor voltage quality index in the future time period Monitoring points for early warning;

The data storage unit is used to store the real-time voltage data and voltage statistical data of the monitoring point, the online prediction model parameters of the future trend of the voltage quality index, the value of each voltage quality index of the monitoring point in each historical time period, and the value of the monitoring point in the future time period. Predicted value of voltage quality index.

The historical time period voltage quality index analysis unit includes a historical time period voltage quality index value calculation module and a historical time period voltage quality index analysis display and alarm module;

The voltage quality index value calculation module in the historical time period is used to select the monitoring point to be analyzed, the voltage quality index and the analysis method to be analyzed, select a specific historical time period according to the selected analysis method, and then calculate the selected monitoring point in the historical time period. The voltage quality index value of the segment;

The voltage quality index analysis display and alarm module in the historical time period is used to display the voltage quality index value of the selected monitoring point in the form of table, curve and bar graph; according to the voltage quality index value of the selected monitoring point and the voltage quality index setting of each level Set the value to determine the monitoring points with poor voltage quality indicators to alarm.

The voltage quality index prediction unit in the future time period includes a monitoring point historical data supplement module, a time series model building module, a gray model building module, a combined forecast model building module, a BP neural network model building module, a voltage quality index prediction module, and a voltage quality index prediction module. Result display and early warning module;

The monitoring point historical data supplement module is used to select the monitoring points that need to be predicted, the voltage quality indicators that need to be predicted, and the prediction method, and select a specific future time period according to the selected prediction method. When the index data is missing, the missing voltage quality index data will be supplemented;

The time series model building module is used to establish a time series model that takes the voltage quality index value in the time range from the current moment to a certain historical moment as input, and outputs the voltage quality index value in the future time period to be predicted;

The gray model establishment module is used to establish a gray model that takes the voltage quality index value in the time range from the current moment to a certain historical moment as input, and outputs the voltage quality index value in the future time period to be predicted;

The combined prediction model building module is used to establish the voltage quality index value of the future time period predicted by the time series model and the voltage quality index value of the future time period predicted by the gray model. Input, the voltage quality of the future time period predicted by the time series model The weighted sum of the index value and the voltage quality index value of the future time period predicted by the gray model is the output combined prediction model;

The BP neural network model building module is used to input the voltage quality index value of the future time period predicted by the combined forecasting model, and output the BP neural network model with the error of the voltage quality index value of the future time period predicted by the combined forecasting model;

The voltage quality index prediction module is used to establish the error of the voltage quality index value of the future time period predicted by the combination prediction model and the voltage quality index value of the future time period predicted by the BP neural network model output as input, and the combined prediction The sum of the error sum of the voltage quality index value predicted by the model in the future time period and the combined prediction model output by the BP neural network model is the future trend online prediction model of the output voltage quality index, and according to The established online prediction model of the future trend of the voltage quality index predicts the future trend of the voltage quality index of the selected monitoring point, and obtains the prediction result of the voltage quality index;

The voltage quality index prediction result display and early warning module is used to display the online prediction results of the future trend of the voltage quality index of the selected monitoring point in the form of tables, curves and bar graphs, and according to the online prediction of the future trend of the voltage quality index of the selected monitoring point Prediction results, the set values of voltage quality indicators at each level, determine the monitoring points of poor voltage quality indicators in the future, and carry out early warning.

In this embodiment, the voltage monitoring method for improving the voltage quality of the power grid has a process as shown in Figure 2, including the following steps:

Step 1: Obtain real-time voltage data and voltage statistical data of multiple monitoring points;

Real-time voltage data includes 5-minute real-time voltage data;

Voltage statistical data include daily voltage statistical data, monthly voltage statistical data, quarterly voltage statistical data, annual voltage statistical indicators and indicator statistical data of departments at all levels at monitoring points;

Among them, the monthly statistical data of voltage includes the statistical value of the voltage of the current month, the statistical value of the voltage of the previous month, the typical day and time data of the current month, the typical day and time data of the last month, the index statistical data of the electricity department at all levels in the current month, the reliability data of the voltage reliability of the last month Data, statistical values of power outages in the current month and last month's power outages; voltage annual statistical indicators include voltage annual pass rate, voltage annual over-limit rate and voltage annual over-low limit rate; indicator statistics of departments at all levels include daily pass rates of departments at all levels Statistical data, monthly qualification rate statistics, quarterly qualification rate statistics and annual qualification rate statistics; departments at all levels include substations, power supply stations, sub-control districts, county power supply bureaus and regional power bureaus;

Step 2: According to the voltage statistical data collected at each monitoring point, analyze the historical situation of the voltage quality index, as shown in Figure 7;

Step 2-1: Select the monitoring points to be analyzed;

The methods of selecting monitoring points to be analyzed include selecting a single monitoring point, selecting multiple monitoring points, selecting all relevant monitoring points of a line at one time, selecting all monitoring points of a substation at one time, and selecting all monitoring points of the county power supply bureau at one time.

Step 2-2: Select the voltage quality index to be analyzed. The voltage quality index includes voltage qualification rate, voltage standard deviation and voltage probability density;

Step 2-3: Select the analysis method of year-by-quarter, year-by-month, quarter-by-month or month-by-day, and select a specific historical time period according to the selected analysis method;

When the analysis method is year by quarter, select the historical time period as year; when the analysis method is year by month, select the historical time period as year; when the analysis method is quarterly by month, select the historical time period as year and quarter; When the analysis mode is month by day, select the historical time period as year and month.

Step 2-4: Calculate the voltage quality index value of the selected monitoring point in the historical time period according to the selected analysis method and historical time period;

Voltage qualification rate = (1-voltage overrun time/total operation statistics time)*100%;

When s is the daily voltage standard deviation, n is the number of sampling points in one day, is the daily average voltage; when s is the monthly voltage standard deviation, n is the number of days in a month; is the monthly average voltage; when s is the annual voltage standard deviation, n is 12; is the annual average voltage;

Voltage density deviation = ∫(g(x)-h(x))dx, x represents the voltage value, g(x) is the actual voltage probability density distribution, h(x) is the expected voltage distribution obtained from historical voltage data.

Step 2-5: Display the voltage quality index value of the selected monitoring point in the form of table, curve and bar graph;

This embodiment shows the voltage standard deviation of the selected monitoring point in the form of a table, see Table 1:

Table 1 The voltage quality index value of the selected monitoring points

time Monitoring point number Monitoring point name Voltage index name Voltage index value evaluate January 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.32 better February 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.66 better March 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.01 better April 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.16 better May 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.07 better June 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.41 better July 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 2.12 better

August 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 3.22 generally September 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 0.96 better October 2011 02010101010000001 No. 1 LAN monitoring point voltage standard deviation 39.34 Difference

The present embodiment shows the voltage quality index value of the selected monitoring point in the form of a curve as shown in Figure 3, the curve in this figure is the voltage standard deviation of the month in 2011; the voltage of the selected monitoring point is displayed in the form of a bar graph The quality index values are shown in Figure 4.

Step 2-6: According to the voltage quality index value of the selected monitoring point and the setting value of the voltage quality index of each level, determine the monitoring point with poor voltage quality index and send an alarm so that the electric power personnel can monitor the abnormal voltage quality of the monitoring point. diagnostic treatment;

The setting value of the standard deviation index of each grade voltage is: good: [0,1); better: [1,3); general: [3,10); poor: [10,20); poor [20,+ ∞).

According to the above results, it can be seen very clearly that the voltage standard deviation index of the monitoring point in October 2011 was extremely poor, and an alarm display was performed to guide the electric power personnel to check the voltage quality of the monitoring point in October 2011, so as to realize the monitoring of the voltage. In-depth analysis of quality;

Step 3: According to the current collected real-time voltage data of each monitoring point, online prediction is made on the future trend of the voltage quality index, as shown in Figure 8;

Step 3-1: Select the monitoring points that need to be predicted;

The methods of selecting the monitoring points that need to be predicted include selecting a single monitoring point, selecting multiple monitoring points, selecting all relevant monitoring points of the line at one time, selecting all monitoring points of the substation at one time, and selecting all monitoring points of the county power supply bureau at one time.

Step 3-2: Select the voltage quality index that needs to be predicted;

Step 3-3: Select the forecast method of year-by-quarter, year-by-month, quarter-by-month or month-by-day, and select a specific future time period according to the selected forecast method;

When the forecast method is year-by-quarter, the time of the forecast result is the remaining quarter of the year; when the monitoring method is year-by-month, the time of the forecast result is the remaining month of the year; The time of the result is the remaining month of the quarter; when the monitoring method is month-by-day, the time of the forecast result is the remaining days of the month;

Step 3-4: When the voltage quality index data of the selected monitoring point is missing in the historical time period, supplement the missing voltage quality index data;

The method used for data supplementation is the cubic spline interpolation algorithm. The cubic spline interpolation function S(x)∈C ² [a,b], a=x ₀ <x ₁ <...<x _n ＝b, y _j ＝f( x _j ), S(x) is a cubic polynomial on each small interval [x _j ,x _j+1 ]: S _j (x)=a _j x ³ +b _j x ² +c _j x+d _j j ＝0,1,...,n-1

Where a _j , b _j , c _j , d _j are to be determined, and make it satisfy:

S(x _j )=y _j j=0,1,…,n

$\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}{\mathrm{<span\; class="notranslate">\; lim</span>}}\mathrm{<span\; class="notranslate">\; S</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; S</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}$

$\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}{\mathrm{<span\; class="notranslate">\; lim</span>}}{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}$

$\underset{\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; \&Right\; Arrow;</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}}{\mathrm{<span\; class="notranslate">\; lim</span>}}{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; S</span>}}^{<span\; class="notranslate">\; \&prime;</span><span\; class="notranslate">\; \&prime;</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}$

Where x _j represents the jth node in the interval [a,b], y _j represents the function value corresponding to node x _j , S(x _j ) represents the function value of the cubic spline interpolation function corresponding to node x _j , Indicates the limit value of S(x) at node x _j , Indicates the limit value of the first derivative of S(x) at node x _j , Indicates the limit value of the second derivative of S(x) at node x _j , S'(x _j ) indicates the first derivative of S(x) at node x _j , S″(x _j ) indicates S(x) at node x _j Second Derivative.

Step 3-5: Establish an online prediction model of the future trend of the voltage quality index. The input of the model is the value of the voltage quality index within the time range from the current moment to a certain historical moment, and the output of the model is the future to be predicted The voltage quality index value of the time period, as shown in Figure 9;

Step 3-5-1: Establish a time series model. The input of the time series model is the voltage quality index value within the time range from the current moment to a certain historical moment, and the output of the time series model is the future time period to be predicted. Voltage quality index value;

In this embodiment, the time series model adopts the ARIMA (p, d, q) model, where p, d, and q respectively represent the order of the autoregressive model, the order of the difference, and the order of the moving average model. Using a mathematical model to approximate this sequence, its representation is:

Among them, X _t is an unknown variable, is the hysteresis operator, a _j represents the autoregressive model parameter coefficient, b _j represents the moving average model parameter, ε _t represents the sequence of independent and identically distributed random variables.

Step 3-5-2: Establish a gray model, the input of the gray model is the voltage quality index value in the time range from the current moment to a certain historical moment, and the output of the gray model is the voltage quality in the future time period to be predicted Index value;

(1) Obtain X ^{( 1 ) =} [ ^{x (1)} (1),x ⁽¹⁾ (2),…,x ⁽¹⁾ (n)], where x ⁽⁰⁾ (i)＞0 (i=1,2,…,n) means voltage quality Index value; $x^{\left(1\right)}\left(t\right)=\underset{k=1}{\overset{t}{\mathrm{\&Sigma;}}}{x}^{\left(0\right)}\left(k\right),t=\mathrm{1,2},...,\mathrm{no};$

(2) Construct accumulation matrix B and constant item vector Y _n , namely

$\mathrm{<span\; class="notranslate">\; B</span>}<span\; class="notranslate">\; =</span>\left[\begin{array}{cc}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>& \mathrm{<span\; class="notranslate">\; 1</span>}\\ <span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 3</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>& \mathrm{<span\; class="notranslate">\; 1</span>}\\ <span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span>& \mathrm{<span\; class="notranslate">\; 1</span>}\\ <span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>& \mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right]<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; .</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ]</span>}^{\mathrm{<span\; class="notranslate">\; T</span>}}$

(3) Solve the gray parameters with the least squares method:

(4) The whitening differential equation corresponding to the GM(1,1) model is:

(5) Substitute the gray parameters into the time corresponding function: ${\hat{x}}^{\left(1\right)}(t+1)=(x^{\left(0\right)}\left(1\right)-\frac{u}{a})e^{-\mathrm{at}}+\frac{u}{a},t=\mathrm{1,2},...,\mathrm{no},$ Accumulated and restored to get ${\hat{x}}^{\left(0\right)}(t+1)={\hat{x}}^{\left(1\right)}(t+1)-{\hat{x}}^{\left(1\right)}\left(t\right);$

(6) yes ${\hat{x}}^{\left(1\right)}(t+1)=(x^{\left(0\right)}\left(1\right)-\frac{u}{a})e^{-\mathrm{at}}+\frac{u}{a},t=\mathrm{1,2},...,\mathrm{no}$ Reduction and derivation get the prediction equation:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; t</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; a</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; x</span>}}^{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 0</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; u</span>}}{\mathrm{<span\; class="notranslate">\; a</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; at</span>}}$

Step 3-5-3: Establish a combined forecasting model, the input of the combined forecasting model includes the voltage quality index value of the future time period predicted by the time series model and the voltage quality index value of the future time period predicted by the gray model, and the output of the combined forecasting model It is the weighted sum of the voltage quality index value of the future time period predicted by the time series model and the voltage quality index value of the future time period predicted by the gray model, that is, the voltage quality index value of the future time period predicted by the combined forecasting model;

Combined forecasting model:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; x</span>}}}_{\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; ,</span>$

Where t=1,2,...,n, the actual observed value of the voltage quality index in period t is x _t , w ₁ is the weight coefficient of the time series model, w ₂ is the weight coefficient of the gray model, is the predicted value of the voltage quality index of the t-th period time series model, is the predicted value of the voltage quality index of the gray model in period t. The determination of the weight coefficient is as follows:

$\left\{\begin{array}{c}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}<span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; \&times;</span><span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; |</span>& \mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; \&le;</span>\mathrm{<span\; class="notranslate">\; 10</span>}\\ \frac{{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}{\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; e</span>}}_{\mathrm{<span\; class="notranslate">\; j</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}& \mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ></span>\mathrm{<span\; class="notranslate">\; 10</span>}\end{array}\right.\\ {\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right.$

In the formula, e _j is the error mean square error of the jth model (j=1 represents the time series model, j=2 represents the gray model), namely $e_j=\frac{1}{\mathrm{no}}\underset{i=1}{\overset{\mathrm{no}}{\mathrm{\&Sigma;}}}{(x_j-{\hat{x}}_j)}^2.$

Step 3-5-4: Establish a BP neural network model. The input of the BP neural network model is the voltage quality index value of the future time period predicted by the combined forecasting model, and the output of the BP neural network model is the value of the future time period predicted by the combined forecasting model. The error of the voltage quality index value;

(1) Determining the number of network layers: a three-layer network with a single hidden layer;

(2) Determine the number of neurons in each layer of the network: select the number of neurons in the input layer to be 4, the number of neurons in the output layer to be 1, and set the initial number of nodes in the hidden layer to be 2;

(3) Sample selection and data preprocessing: group the obtained voltage quality index data into groups, one set of voltage quality index data is used to form training samples, and the other set of voltage quality index data is used to form test samples. Large differences lead to large network prediction errors, and it is necessary to normalize the input voltage quality index sample data: Where X _t is the original sample data, X _max and X _min are the maximum and minimum values of the original variable X _t respectively; S _t is the transformed value of X _t ;

Model training: take the hidden layer activation function as a logarithmic sigmoid function, the output layer activation function as a pure linear function, select the training function as the momentum gradient descent and adaptive learning rate training function, and the learning function as the momentum gradient descent learning function. Use the training samples to follow the steps of the neural network algorithm to train the network repeatedly until the network converges to a certain standard. Otherwise, repeatedly change the initial weights of the network, or even the topology of the network, until the training result is satisfactory, that is, the error of the voltage quality index value in the future time period predicted by the combined prediction model is obtained.

Step 3-5-5: Establish the future trend online prediction model of the voltage quality index, the input of the model includes the voltage quality index value of the future time period predicted by the combined prediction model and the future time predicted by the combined prediction model output by the BP neural network model The error of the voltage quality index value of the segment, the output of the online prediction model is the error of the voltage quality index value of the future time period predicted by the combined prediction model and the voltage quality index value of the future time period predicted by the combined prediction model output by the BP neural network model The sum is the prediction result of the voltage quality index.

Predict the sample data of the voltage quality index with the weighted combination prediction model, and obtain the predicted value of the voltage quality index and the prediction error e of the voltage quality index, and use the neural network model to obtain the predicted value of the prediction deviation e of the voltage quality index The prediction result of the final voltage quality index is obtained.

Step 3-6: Predict the future trend of the voltage quality index of the selected monitoring point according to the established online prediction model of the future trend of the voltage quality index;

Step 3-7: Display the online prediction results of the future trend of the voltage quality indicators of the selected monitoring points in the form of tables, curves and bar graphs;

This implementation mode displays the voltage qualification rate of the selected monitoring point in the form of a table, as shown in Table 2:

Table 2 The voltage qualification rate of the selected monitoring points

time Monitoring point number Monitoring point name Voltage index name Voltage index value evaluate January 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 96.88% generally February 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 96.78% generally March 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 97.08% generally April 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 97.98% generally May 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 96.58% generally June 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 97.98% generally July 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 97.98% generally August 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 96.58% generally September 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 94.62% generally October 2011 02010101010000002 No. 2 LAN monitoring point Voltage pass rate 98.47% better January 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 97.08% generally February 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 96.88% generally March 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 96.78% generally April 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 96.58% generally May 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 96.58% generally June 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 97.98% generally August 2011 02010101010000004 No. 4 LAN monitoring point Voltage pass rate 70.56% Difference

Step 3-8: According to the online prediction results of the future trend of the voltage quality indicators of the selected monitoring points and the set values of the voltage quality indicators of each level, determine the monitoring points with poor voltage quality indicators at the future time, and give early warning, so that the electric power personnel can monitor the existing voltage The monitoring points of hidden quality problems are dealt with in advance.

The setting values of the pass rate indicators for each grade are: good: (99.5%, 100%]; better: (98.0%, 99.5%]; general: (90.0%, 98.0%]; poor: (80.0%, 90.0 %]; difference (0,90.0%].

According to the above display results, it can be clearly seen that the voltage qualification rate index of the No. 4 LAN monitoring point in August 2011 was extremely poor, and the early warning display was carried out to guide the electric power personnel to check the voltage quality of the No. 4 LAN monitoring point in advance. , which improves the effectiveness of grid voltage quality monitoring.

Claims (10)

1. a voltage monitoring method that improves grid voltage quality, is characterized in that, comprises the following steps:

Step 1: real-time voltage data and the voltage statistic data of obtaining a plurality of monitoring points;

Described voltage statistic data comprise voltage day statistics, voltage moon statistics, voltage season statistics, the voltage year statistical indicator and the indicator-specific statistics data of departments at different levels of monitoring point;

Data when wherein, the voltage moon, statistics comprised of that month voltage statistic value, last month voltage statistic value, of that month typical case day, last month data during typical case day, of that month voltage reliability data, last month voltage reliability data, of that month power-failure counting value and last month power-failure counting value; The annual statistical indicator of voltage comprises voltage year qualification rate, the annual super upper limit rate of voltage and the annual super lower limit rate of voltage; The indicator-specific statistics data of departments at different levels comprise day qualification rate statistics, month qualification rate statistics, the season qualification rate statistics and year qualification rate statistics of departments at different levels;

Step 2: according to the voltage statistic data of each monitoring point collecting, the account of the history of quality of voltage index is analyzed;

Step 2-1: the monitoring point of selecting to need analysis;

Step 2-2: select to need the quality of voltage index of analysis, quality of voltage index comprises rate of qualified voltage, voltage standard is poor and voltage probability density;

Step 2-3: select year quarterly, year monthly, season monthly or the moon analysis mode per diem, according to selected analysis mode, select concrete historical time section;

Step 2-4: calculate selected monitoring point in the quality of voltage desired value of this historical time section according to selected analysis mode and historical time section;

Step 2-5: the quality of voltage desired value that shows selected monitoring point with the form of form, curve and excellent figure;

Step 2-6: according to the quality of voltage desired value of selected monitoring point, each voltage gradation quality index set point, determine the monitoring point of quality of voltage index error, report to the police, so that electric power personnel carry out diagnostic process to the quality of voltage abnormal conditions of this monitoring point;

Step 3: according to the real-time voltage data of current each monitoring point collecting, the future trend of quality of voltage index is carried out to on-line prediction;

Step 3-1: the monitoring point of selecting to need prediction;

Step 3-2: the quality of voltage index of selecting to need prediction;

Step 3-3: select year quarterly, year monthly, season monthly or the moon prediction mode per diem, according to the prediction mode of selecting, select concrete future time section;

Step 3-4: when the quality of voltage achievement data of selected monitoring point in historical time section exists disappearance, the quality of voltage achievement data of disappearance is supplemented;

Step 3-5: the future trend on-line prediction model of setting up quality of voltage index, this model be input as the quality of voltage desired value in the scope during this period of time from current time to certain historical juncture, this model is output as the quality of voltage desired value of the future time section that will predict;

Step 3-6: the future trend of the quality of voltage index of selected monitoring point is predicted according to the future trend on-line prediction model of the quality of voltage index of setting up;

Step 3-7: the on-line prediction result that shows the quality of voltage index future trend of selected monitoring point with the form of form, curve and excellent figure;

Step 3-8: according to the on-line prediction result of the quality of voltage index future trend of selected monitoring point, each voltage gradation quality index set point, determine the following monitoring point of quality of voltage index error constantly, carry out early warning, so that electric power personnel are to existing the monitoring point of quality of voltage potential problem to carry out advanced processing.

2. the voltage monitoring method of raising grid voltage quality according to claim 1, is characterized in that: the future trend on-line prediction model of setting up quality of voltage index in step 3-5 carries out as follows:

Step 3-5-1: settling time series model, time series models be input as the quality of voltage desired value in the scope during this period of time from current time to certain historical juncture, time series models are output as the quality of voltage desired value of the future time section that will predict;

Step 3-5-2: set up gray model, gray model be input as the quality of voltage desired value in the scope during this period of time from current time to certain historical juncture, gray model output is the quality of voltage desired value of the future time section that will predict;

Step 3-5-3: set up combination forecasting, the input of combination forecasting comprises the quality of voltage desired value of future time section of time series models prediction and the quality of voltage desired value of the future time section of Grey Model, combination forecasting is output as the weighting sum of the quality of voltage desired value of future time section of time series models prediction and the quality of voltage desired value of the future time section of Grey Model, i.e. the quality of voltage desired value of the future time section of combination forecasting prediction;

Step 3-5-4: set up BP neural network model, the quality of voltage desired value of the future time section that is input as combination forecasting prediction of BP neural network model, BP neural network model is output as the error of quality of voltage desired value of the future time section of combination forecasting prediction;

Step 3-5-5: the future trend on-line prediction model of setting up quality of voltage index, the input of this model comprises the error of the quality of voltage desired value of the quality of voltage desired value of future time section of combination forecasting prediction and the future time section of the combination forecasting prediction of BP neural network model output, on-line prediction model is output as the error sum of the quality of voltage desired value of the quality of voltage desired value of future time section of combination forecasting prediction and the future time section of the combination forecasting prediction of BP neural network model output, be predicting the outcome of quality of voltage index.

3. the voltage monitoring method of raising grid voltage quality according to claim 1, is characterized in that: the mode of the monitoring point that the selection described in the mode of the monitoring point that the selection described in step 2-1 need to be analyzed and step 3-1 need to be predicted includes and selects single monitoring point, selects a plurality of monitoring points, once selects all relevant monitoring points of circuit, once selects all monitoring points of transformer station and once select all monitoring points of county power supply administration.

4. the voltage monitoring method of raising grid voltage quality according to claim 1, is characterized in that: described quality of voltage index comprises rate of qualified voltage, voltage standard is poor and voltage probability density.

5. the voltage monitoring method of raising grid voltage quality according to claim 1, is characterized in that: described in step 2-3, according to selected analysis mode, select concrete historical time section, when analysis mode is year quarterly time, historical time section is chosen as year; When analysis mode is year monthly time, historical time section is chosen as year; When analysis mode is season monthly time, historical time section is chosen as time and season; When analysis mode is the moon per diem time, historical time section is chosen as time and month.

6. the voltage monitoring method of raising grid voltage quality according to claim 1, it is characterized in that: the prediction mode according to selecting described in step 3-3 is selected concrete future time section, when prediction mode is year quarterly time, the residue season that the time under predicting the outcome is the current year; When monitoring mode is year monthly time, the residue month that the time under predicting the outcome is the current year; When monitoring mode is season monthly time, the residue month that the time under predicting the outcome is current season; When monitoring mode is the moon per diem time, the residue number of days that the time under predicting the outcome is this month.

7. the voltage monitoring method of raising grid voltage quality according to claim 1, is characterized in that: the quality of voltage achievement data to disappearance described in step 3-4 supplements the cubic spline interpolation method that adopts.

8. the grid voltage quality monitoring system that realizes the monitoring method of raising grid voltage quality claimed in claim 1, is characterized in that: comprise data capture unit, historical time section quality of voltage index analysis unit, future time section quality of voltage index prediction unit and data storage cell;

Data capture unit is for obtaining real-time voltage data and the voltage statistic data of monitoring point;

Historical time section quality of voltage index analysis unit is for the statistics of each monitoring point of obtaining according to data capture unit, account of the history to quality of voltage index is analyzed, and comprising: select the monitoring point of needs analysis and the quality of voltage index that needs are analyzed; Select year quarterly, year monthly, season monthly or the moon analysis mode per diem select concrete historical time section according to selected analysis mode; According to selected analysis mode and historical time section, calculate selected monitoring point in the quality of voltage desired value of this historical time section; The quality of voltage desired value that shows selected monitoring point with the form of form, curve and excellent figure; According to the quality of voltage desired value of selected monitoring point, each voltage gradation quality index set point, determine to report to the police in the monitoring point of quality of voltage index error;

Future time section quality of voltage index prediction unit is used for according to the real-time voltage data of current each monitoring point obtaining, the future trend of quality of voltage index is carried out to on-line prediction, comprising: select the monitoring point of needs prediction and the quality of voltage index of needs prediction; Select year quarterly, year monthly, season monthly or the moon prediction mode per diem select concrete future time section according to the prediction mode of selecting; When the quality of voltage achievement data of selected monitoring point in historical time section exists disappearance, the quality of voltage achievement data of disappearance is supplemented; Set up the future trend on-line prediction model of quality of voltage index, this model be input as the quality of voltage desired value in the scope during this period of time from current time to certain historical juncture, this model is output as the quality of voltage desired value of the future time section that will predict; According to the future trend on-line prediction model of the quality of voltage index of setting up, the future trend of the quality of voltage index of selected monitoring point is predicted; The on-line prediction result that shows the quality of voltage index future trend of selected monitoring point with the form of form, curve and excellent figure; According to the on-line prediction result of the quality of voltage index future trend of selected monitoring point, each voltage gradation quality index set point, determine that early warning is carried out in the monitoring point of future time section quality of voltage index error;

Data storage cell for the future trend on-line prediction model parameter, monitoring point of storing the real-time voltage data of monitoring point and voltage statistic data, quality of voltage index in every quality of voltage desired value of each historical time section and monitoring point the quality of voltage index prediction value in future time section.

9. the monitoring system of raising grid voltage quality according to claim 8, is characterized in that: described historical time section quality of voltage index analysis unit comprises that historical time section quality of voltage desired value computing module and historical time section quality of voltage index analysis show and alarm module;

Historical time section quality of voltage desired value computing module is for selecting to need the monitoring point of analyzing, quality of voltage index and the analysis mode that needs analysis, according to selected analysis mode, select concrete historical time section, and then calculate selected monitoring point in the quality of voltage desired value of this historical time section;

The demonstration of historical time section quality of voltage index analysis and alarm module are for showing the quality of voltage desired value of selected monitoring point with the form of form, curve and excellent figure; According to the quality of voltage desired value of selected monitoring point, each voltage gradation quality index set point, determine to report to the police in the monitoring point of quality of voltage index error.

10. the monitoring system of raising grid voltage quality according to claim 8, is characterized in that: described future time section quality of voltage index prediction unit comprises that monitoring point historical data complementary module, time series models set up module, gray model and set up module, combination forecasting and set up module, BP Establishment of Neural Model module, quality of voltage index prediction module, quality of voltage index prediction result and show and warning module;

Monitoring point historical data complementary module is for selecting to need the monitoring point of prediction, quality of voltage index and the prediction mode that need to predict, according to the prediction mode of selecting, select concrete future time section, when the quality of voltage achievement data of selected monitoring point in historical time section exists disappearance, the quality of voltage achievement data of disappearance is supplemented;

It is input for the quality of voltage desired value of setting up in the scope during this period of time of take from current time to certain historical juncture that time series models are set up module, and the quality of voltage desired value of the future time section that will predict of take is the time series models of output;

It is the gray model of output for the quality of voltage desired value of setting up in the scope during this period of time of take from current time to certain historical juncture as inputting, take the quality of voltage desired value of the future time section that will predict that gray model is set up module;

Combination forecasting is set up module, and for setting up, to take the quality of voltage desired value of future time section of time series models predictions and the quality of voltage desired value of the future time section of Grey Model be that input, the quality of voltage desired value of future time section of the time series models of take prediction and the weighting sum of the quality of voltage desired value of the future time section of Grey Model are the combination forecasting of exporting;

BP Establishment of Neural Model module is that the error of quality of voltage desired value of the future time section of input, the combination forecasting of take prediction is the BP neural network model of output for take the quality of voltage desired value of future time section of combination forecasting prediction;

For setting up, to take the error of quality of voltage desired value of the quality of voltage desired value of future time section of combination forecasting prediction and the future time section of the combination forecasting prediction of BP neural network model output be input to quality of voltage index prediction module, the future trend on-line prediction model that the error sum of the quality of voltage desired value of the future time section of the combination forecasting prediction of the quality of voltage desired value of the future time section of the combination forecasting of take prediction and the output of BP neural network model is the quality of voltage index of output, and according to the future trend on-line prediction model of the quality of voltage index of setting up, the future trend of the quality of voltage index of selected monitoring point is predicted, obtain predicting the outcome of quality of voltage index,

The demonstration of quality of voltage index prediction result and warning module are for showing the on-line prediction result of the quality of voltage index future trend of selected monitoring point with the form of form, curve and excellent figure, and according to the on-line prediction result of the quality of voltage index future trend of selected monitoring point, each voltage gradation quality index set point, determine the following monitoring point of quality of voltage index error constantly, carry out early warning.