CN104134096A - Intelligent load prediction method based on province and prefecture integration - Google Patents

Abstract

The invention discloses an intelligent load prediction method based on province and prefecture integration. The method sequentially comprises the following steps that: real-time electricity consumption quantity of each city and user electricity consumption information are collected; the actual load and predicted load historical data of the whole network and each city in a sample time period in recent time is collected; an average proportion coefficient of each city at the moment t in the sample time period is calculated; the load and meteorological correlation degree quantification calculation is carried out; according to the electricity consumption quantity, the user electricity consumption information, meteorological indexes and the correlation coefficient absolute value, each load prediction model and each load prediction value are obtained; and according to the load prediction value and the load proportion prediction value of each prefecture-city company, the whole network load prediction value is obtained. The intelligent load prediction method provided by the invention has the advantages that the load prediction result in each prefecture city is sufficiently utilized; a dispatching department can favorably, deeply and finely master a characteristic change rule of load ingredients of each region of a power network; the load prediction graded management is enhanced; and the scientization and fining level of the load prediction can be comprehensively improved.

Description

Based on economizing ground integrated intelligent load forecast method

Technical field

The present invention relates to electric power load monitoring and control technology, be specifically related to a kind of based on economizing ground integrated intelligent load forecast method.

Background technology

Load Prediction In Power Systems refers to from known economy, social development and electricity needs situation, by to the analysis of historical data and research, explore inner link and development and change rule between things, following electricity needs is made to estimation and supposition in advance.Load prediction is in fact the prediction to electricity market demand, is the prerequisite that realizes power grid security, economical operation.Short-term load forecasting is the important component part of load prediction; it is mainly used in forecasting the electric load in coming few minutes, several hours or several weeks; for scheduling, arrange startup-shutdown plan, for electric power application, have great significance as unit commitment, economic load dispatching, optimal load flow.If there is larger error in load prediction, may make electric system need to strengthen start capacity, there is extra system reserve, cause waste economically.

In different seasons, the people's production is different with habits and customs, and the regularity difference of utilization of power has formed the calendar variation of load curve; The geography of different regions, climatic characteristic and power structure have otherness, and are embodied directly on the Morphological Diversity of prefectures and cities' load curve.These othernesses and diversity have increased the difficulty of grasping each province and prefectures and cities' load variations regularity and carrying out accurately predicting.In this case, make full use of by time the impact of comprehensive meteorological index on load fluctuation, to improving the load prediction precision of regional power network, there is positive effect.

Summary of the invention

The technical problem to be solved in the present invention is to provide a kind of based on economizing ground integrated intelligent load forecast method, to improve provincial Utilities Electric Co. short-term load forecasting precision.

For solving the problems of the technologies described above, the present invention by the following technical solutions:

Based on economizing a ground integrated intelligent load forecast method, its step is as follows:

(1) gather real-time prefectures and cities' power consumption and user power utilization information;

(2) gather actual load and the prediction load historical data of interior the whole network of a recent sample period and each districts and cities;

(3) calculate in this sample period each districts and cities at the average proportions coefficient of t moment point, the scale-up factor with to be predicted day each districts and cities of exponential smoothing performance prediction in t moment point, forms each districts and cities at the scale-up factor matrix of t moment point;

Wherein onset index smoothing method model is:

$C_t^i=\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\λ}}_j{C}_{t,j}^i,i=\mathrm{1,2},...,N$

$\left\{\begin{array}{c}\underset{j=1}{\overset{n}{\mathrm{\Σ}}}{\mathrm{\λ}}_j=1\\ 0\≤{\mathrm{\λ}}_j\≤1\end{array}\right.,j=\mathrm{1,2},...,n$

In formula: represent that districts and cities i accounts for the predicted value of the whole network system loading ratio at moment t, before representing, the districts and cities i of j days accounts for the actual value of the whole network system loading ratio at moment t; λ _jrepresent weight coefficient, λ _j=λ (1-λ) ^j-1, λ is constant, and 0< λ <1; N is the number of days of sample period;

(4) obtain the comprehensive meteorological index data of each historical day and to be measured day;

(5) load and quantize to calculate with meteorological degree of correlation, by the absolute value of related coefficient | r _xy| maximum load sequential value and meteorological index sequential value are determined leading meteorological effect factor,

$r_{\mathrm{xy}}=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}(x_i-\stackrel{\&OverBar;}{x})(y_i-\stackrel{\&OverBar;}{y})}{{\left[\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(x_i-\stackrel{\&OverBar;}{x})}^2\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(y_i-\stackrel{\&OverBar;}{y})}^2\right]}^{\frac{1}{2}}}$

Wherein, x _ifor load sequential value, for the load mean value in time interval; y _ifor meteorological index sequential value, for the meteorological index mean value in time interval;

(6), according to described power consumption, user power utilization information, meteorological index and related coefficient absolute value, obtain prefectures and cities' company's load forecasting model and load prediction value;

(7) according to the predicted value of prefectures and cities company load prediction value, described load proportion, obtain the whole network load prediction value.

Comprehensive meteorological index described in described step (4) comprises effective temperature, comfort index, coldness index and Body Comfort Index.

Described effective temperature adopts following formula to calculate:

$T_e=37-\frac{(37-T_a)}{[0.68-0.14R_h+1/(1.76+1.4V^{0.75})]}-0.29T_a(1-R_h)$

In formula, Te represents effective temperature, and Ta represents temperature, R _hrepresent relative humidity, V represents wind speed;

Described humiture comfort index adopts following formula to calculate:

E _T＝T _d-0.55(1-R _d)(58-T _d)

E in formula _tstatement comfort index, Td represents the temperature of 14 o'clock, Rd represents the relative humidity of 14 o'clock;

Described coldness index adopts following formula to calculate:

$H=\mathrm{\&Delta;t}(9.0+10.9\sqrt{V}-V)$

Wherein, H represents coldness index, and V represents wind speed, and Δ T represents the poor of body temperature and temperature around;

Described Body Comfort Index adopts following formula to calculate:

$k=1.8T_a-0.55(1-R_h)-3.2\sqrt{V}+32$

Wherein, k represents Body Comfort Index, and Ta represents temperature, and Rh represents relative humidity, and V represents wind speed.

The advantage of this programme is: 1, invention changes and to this influential meteorologic factor situation, following power load predicted from known electric load.The present invention take full advantage of by time the impact of comprehensive meteorological index on load fluctuation, improved the load prediction precision of regional power network, contribute to grid company more arranged rational, distribute electric power supply ability rationally, effectively implement ordered electric, reduce operation of power networks risk, guarantee to realize the maximization of enterprise, social benefit.2, the real-time distributed generating gathering according to intelligent terminal and user power utilization information, again in conjunction with the historical power information of enterprise and other basic data, carry out information analysis and computing based on model, Dynamic Generation enterprise controllable burden inventory, load control sequence etc., formulate distributed power generation plan and user power utilization plan, for Utilities Electric Co.'s load control plan, formulate main reference is provided, distribution dispatching center carries out the accurately predicting of distributed power generation amount and load power consumption according to these information, for power distribution network provides strategy and the foundation of Optimized Operation.

Embodiment

For making the object, technical solutions and advantages of the present invention clearer, below the present invention is described in further detail.

Embodiments of the present invention relate to a kind of based on economizing ground integrated intelligent load forecast method, comprise the following steps:

(1) gather real-time distributed power generation amount and user power utilization information;

(2) gather actual load and the prediction load historical data of interior the whole network of a recent sample period and each districts and cities;

(3) calculate in this sample period each districts and cities at the average proportions coefficient of t moment point, the scale-up factor with to be predicted day each districts and cities of exponential smoothing performance prediction in t moment point, forms each districts and cities at the scale-up factor matrix of t moment point;

Wherein onset index smoothing method model is:

$C_t^i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_j{C}_{t,j}^i,i=\mathrm{1,2},...,N$

$\left\{\begin{array}{c}\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_j=1\\ 0\&le;{\mathrm{\&lambda;}}_j\&le;1\end{array}\right.,j=\mathrm{1,2},...,n$

In formula: represent that districts and cities i accounts for the predicted value of the whole network system loading ratio at moment t, before representing, the districts and cities i of j days accounts for the actual value of the whole network system loading ratio at moment t; λ _jrepresent weight coefficient, λ _j=λ (1-λ) ^j-1, λ is constant, and 0< λ <1; N is the number of days of sample period;

(4) obtain the comprehensive meteorological index data of each historical day and to be measured day, described comprehensive meteorological index comprises effective temperature, comfort index, coldness index and Body Comfort Index;

Described effective temperature is that human body is under different temperature, humidity and wind friction velocity, the hotness index producing, effective temperature is with static saturated atmosphere (wind speed=0 o'clock, relative humidity=100%) under condition, make people produce comfortable temperature, represent the same sensation that different wind speed, different relative humidity, different temperature produce people.For example following three kinds of said sensations of situation are all equivalent to 17.7 ℃ of true feeling temperature.

1. temperature is 17.7 ℃, relative humidity 100, wind speed 0;

2. temperature is 22.4 ℃, relative humidity 75%, wind speed 0.5m/s;

3. temperature is 25 ℃, relative humidity 20%, wind speed 2.5m/s.

Its computing formula is:

$T_e=37-\frac{(37-T_a)}{[0.68-0.14R_h+1/(1.76+1.4V^{0.75})]}-0.29T_a(1-R_h)---\left(1\right)$

In formula, Te represents effective temperature, and Ta represents temperature, R _hrepresent relative humidity, V represents wind speed;

From formula (1): 1. wind speed increases effective temperature is reduced, but along with temperature increases, wind speed is more and more less on the impact of effective temperature, and when temperature surpasses 37 ℃, wind speed increases makes effective temperature raise on the contrary.2. the affect more complicated of humidity on effective temperature, finds by calculating: in the situation that wind speed is certain, when temperature is higher, humidity increase increases effective temperature; When temperature is lower, humidity increase makes effective temperature reduce.As work as V=3m/s, time, 15 ℃～17 ℃ of temperature are its critical values, when temperature is greater than 17 ℃, the increase of relative humidity increases effective temperature, and temperature while being less than 15 ℃ humidity increase effective temperature is reduced.This index of effective temperature has reflected actual conditions substantially as can be seen from the above analysis, and the content of the daily weather forecast of parameter Dou Shi observatory of using due to its, convenience of calculation.

Described humiture comfort index adopts following formula to calculate:

E _T＝T _d-0.55(1-R _d)(58-T _d)

E in formula _tstatement comfort index, Td represents the temperature of 14 o'clock, Rd represents the relative humidity of 14 o'clock;

Described coldness index adopts following formula to calculate:

$H=\mathrm{\&Delta;t}(9.0+10.9\sqrt{V}-V)$

Wherein, H represents coldness index, and V represents wind speed, and Δ T represents the poor of body temperature and temperature around;

Described Body Comfort Index adopts following formula to calculate:

$k=1.8T_a-0.55(1-R_h)-3.2\sqrt{V}+32$

Wherein, k represents Body Comfort Index, and Ta represents temperature, and Rh represents relative humidity, and V represents wind speed.

(5) load and quantize to calculate with meteorological degree of correlation, by the absolute value of related coefficient | r _xy| maximum load sequential value and meteorological index sequential value are determined leading meteorological effect factor,

$r_{\mathrm{xy}}=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}(x_i-\stackrel{\&OverBar;}{x})(y_i-\stackrel{\&OverBar;}{y})}{{\left[\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(x_i-\stackrel{\&OverBar;}{x})}^2\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(y_i-\stackrel{\&OverBar;}{y})}^2\right]}^{\frac{1}{2}}}$

Wherein, x _ifor load sequential value, for the load mean value in time interval; y _ifor meteorological index sequential value, for the meteorological index mean value in time interval;

(6), according to described power consumption, user power utilization information, meteorological index and related coefficient absolute value, obtain prefectures and cities' company's load forecasting model and load prediction value;

(7) according to the predicted value of prefectures and cities company load prediction value and load proportion, obtain the whole network load prediction value.

The above embodiment has only expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but can not therefore be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.

Claims (3)

1. based on economizing ground integrated intelligent load forecast method, it is characterized in that, comprise the following steps:

(1) gather real-time prefectures and cities' power consumption and user power utilization information;

(2) gather actual load and the prediction load historical data of interior the whole network of a recent sample period and each districts and cities;

(3) calculate in this sample period each districts and cities at the average proportions coefficient of t moment point, the scale-up factor with to be predicted day each districts and cities of exponential smoothing performance prediction in t moment point, forms each districts and cities at the scale-up factor matrix of t moment point; Wherein onset index smoothing method model is:

$C_t^i=\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_j{C}_{t,j}^i,i=\mathrm{1,2},...,N$

$\left\{\begin{array}{c}\underset{j=1}{\overset{n}{\mathrm{\&Sigma;}}}{\mathrm{\&lambda;}}_j=1\\ 0\&le;{\mathrm{\&lambda;}}_j\&le;1\end{array}\right.j=\mathrm{1,2},...,n$

In formula: represent that districts and cities i accounts for the predicted value of the whole network system loading ratio at moment t, before representing, the districts and cities i of j days accounts for the actual value of the whole network system loading ratio at moment t; λ _jrepresent weight coefficient, λ _j=λ (1-λ) ^j-1, λ is constant, and 0< λ <1; N is the number of days of sample period;

(4) obtain the comprehensive meteorological index data of each historical day and to be measured day;

(5) load and quantize to calculate with meteorological degree of correlation, by the absolute value of related coefficient | r _xy| maximum load sequential value and meteorological index sequential value are determined leading meteorological effect factor,

$r_{\mathrm{xy}}=\frac{\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}(x_i-\stackrel{\&OverBar;}{x})(y_i-\stackrel{\&OverBar;}{y})}{{\left[\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(x_i-\stackrel{\&OverBar;}{x})}^2\underset{i=1}{\overset{n}{\mathrm{\&Sigma;}}}{(y_i-\stackrel{\&OverBar;}{y})}^2\right]}^{\frac{1}{2}}}$

Wherein, x _ifor load sequential value, for the load mean value in time interval; y _ifor meteorological index sequential value, for the meteorological index mean value in time interval;

(6), according to described power consumption, user power utilization information, meteorological index and related coefficient absolute value, obtain prefectures and cities' company's load forecasting model and load prediction value;

(7) according to the predicted value of prefectures and cities company load prediction value and load proportion, obtain the whole network load prediction value.

2. according to claim 1 based on economizing ground integrated intelligent load forecast method, it is characterized in that: the comprehensive meteorological index described in step (4) comprises effective temperature, comfort index, coldness index and Body Comfort Index.

3. according to claim 2 based on economizing ground integrated intelligent load forecast method, it is characterized in that:

Described effective temperature adopts following formula to calculate:

$T_e=37-\frac{(37-T_a)}{[0.68-0.14R_h+1/(1.76+1.4V^{0.75})]}-0.29T_a(1-R_h)$

In formula, Te represents effective temperature, and Ta represents temperature, R _hrepresent relative humidity, V represents wind speed;

Described humiture comfort index adopts following formula to calculate:

E _T＝T _d-0.55(1-R _d)(58-T _d)

E in formula _tstatement comfort index, Td represents the temperature of 14 o'clock, Rd represents the relative humidity of 14 o'clock;

Described coldness index adopts following formula to calculate:

$H=\mathrm{\&Delta;t}(9.0+10.9\sqrt{V}-V)$

Wherein, H represents coldness index, and V represents wind speed, and Δ T represents the poor of body temperature and temperature around;

Described Body Comfort Index adopts following formula to calculate:

$k=1.8T_a-0.55(1-R_h)-3.2\sqrt{V}+32$

Wherein, k represents Body Comfort Index, and Ta represents temperature, and Rh represents relative humidity, and V represents wind speed.