CN117786370B - Information intelligent analysis system for gridding service terminal - Google Patents
Information intelligent analysis system for gridding service terminal Download PDFInfo
- Publication number
- CN117786370B CN117786370B CN202410205940.8A CN202410205940A CN117786370B CN 117786370 B CN117786370 B CN 117786370B CN 202410205940 A CN202410205940 A CN 202410205940A CN 117786370 B CN117786370 B CN 117786370B
- Authority
- CN
- China
- Prior art keywords
- data
- electricity consumption
- value
- degree
- periodic
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000004458 analytical method Methods 0.000 title claims abstract description 23
- 230000005611 electricity Effects 0.000 claims abstract description 189
- 230000000737 periodic effect Effects 0.000 claims abstract description 127
- 238000000034 method Methods 0.000 claims abstract description 58
- 238000004364 calculation method Methods 0.000 claims description 22
- 238000012935 Averaging Methods 0.000 claims description 6
- 238000004590 computer program Methods 0.000 claims description 5
- 238000012216 screening Methods 0.000 claims description 3
- 230000002159 abnormal effect Effects 0.000 abstract description 12
- 238000007405 data analysis Methods 0.000 abstract description 2
- 230000006870 function Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 2
- 238000005265 energy consumption Methods 0.000 description 2
- 238000013507 mapping Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000005856 abnormality Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to the technical field of electric quantity data analysis of gridding service terminals, in particular to an intelligent information analysis system for gridding service terminals. The method screens out the larger value and the smaller value of the electricity consumption, obtains the periodic parameter according to the discrete degree, and divides the electricity consumption data into strong periodic data and weak periodic data; according to the deviation degree and the time sequence position of the data points, a first weight coefficient of the data points in the moving window in the weak periodic data is obtained; combining the time sequence positions of the data points in the moving window and the position difference and the exceeding degree of the latest periodical time points to obtain a second weight coefficient of each data point in the moving window in the strong periodical data; and carrying out predictive analysis on the electricity consumption data according to the weight coefficient. The invention considers the integral characteristics of the data, and if abnormal data occurs in the power consumption data, the weight coefficient of each data point in the moving window can be properly adjusted, so that the prediction result is more accurate.
Description
Technical Field
The invention relates to the technical field of electric quantity data analysis of gridding service terminals, in particular to an intelligent information analysis system for gridding service terminals.
Background
The grid service terminal has an important role in the power industry, can provide real-time electricity consumption data for enterprises, can analyze the electricity consumption data of the enterprises by utilizing historical electricity consumption data, so that customized energy efficiency optimization suggestions and strategies are provided for the enterprises, high-energy-consumption equipment is arranged to operate in electricity consumption valley periods, and low-energy-consumption equipment is arranged to operate in electricity consumption higher periods.
The power consumption of enterprises is time sequence data. In the prior art, a moving average method is often used for predicting the electricity consumption through historical electricity consumption data, but the conventional moving average method does not consider the weight of data in a moving window, or only data with far date is given a lower weight value, data with near date is given a higher weight value, the integral characteristics of the data are not considered, and if abnormal data occur in the electricity consumption data, the prediction result is influenced, so that the deviation of the prediction result is overlarge.
Disclosure of Invention
In order to solve the technical problems that in the prior art, when a moving average method is used for predicting the electricity consumption, the integral characteristics of electricity consumption data are not considered, so that the prediction result is influenced if abnormal data appear, and the deviation of the prediction result is overlarge, the invention aims to provide an information intelligent analysis system for a gridding service terminal, and the adopted technical scheme is as follows:
An intelligent information analysis system for a gridding service terminal, comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the following steps when executing the computer program:
acquiring data of each group of electricity consumption in a preset historical period of each electricity consumption enterprise;
screening each group of electricity consumption data into two types according to the data value, and obtaining an electricity consumption maximum value and an electricity consumption minimum value in each group of electricity consumption data; according to the discrete degree of each of the electricity consumption maximum value and the electricity consumption minimum value, periodic parameters of each group of electricity consumption data are obtained; dividing the electricity consumption data into strong periodic data and weak periodic data according to the periodic parameters;
In the calculation process of a moving average method, according to the deviation degree and time sequence position of each data point in a moving window in weak periodic data, a first weight coefficient of each data point in the moving window in the weak periodic data is obtained; acquiring the cycle time points of the power consumption maximum value and the power consumption minimum value in the strong periodic data; obtaining the exceeding degree of each data point in the strong periodic data on the power consumption larger value and the power consumption smaller value; correcting the first weight coefficient by utilizing the time sequence position of each data point in the moving window and the position difference of the nearest periodic time point and the exceeding degree to obtain a second weight coefficient of each data point in the moving window in strong periodic data;
And according to the first weight coefficient and the second weight coefficient, carrying out predictive analysis on the electricity consumption data by using a moving average method.
Further, the method for obtaining the electricity consumption maximum value and the electricity consumption minimum value in each group of electricity consumption data comprises the following steps:
Carrying out Gaussian distribution model fitting on each group of electricity consumption data, and obtaining the mean value and standard deviation of each group of electricity consumption data according to the Gaussian distribution model;
taking all data points larger than the mean value plus standard deviation as the electricity consumption maximum value of each group of electricity consumption data; all data points less than the mean minus the standard deviation are taken as the power usage bias value for each set of power usage data.
Further, the periodic parameter acquisition method of the electricity consumption data comprises the following steps:
respectively calculating the overall discrete degree of all the electricity consumption maximum values and the electricity consumption minimum values; the integral discrete degree is the sum of the discrete degree of the data value and the discrete degree of the occurrence time interval;
And normalizing and adding the whole discrete degrees of all the power consumption maximum values and the power consumption minimum values respectively to obtain the periodic parameters.
Further, dividing the electricity consumption data into strong periodic data and weak periodic data according to the periodic parameters includes:
A first threshold value is preset, the electricity consumption data with the periodic parameter being larger than the first threshold value is used as strong periodic data, and the electricity consumption data with the periodic parameter being smaller than the first threshold value is used as weak periodic data.
Further, the method for acquiring the cycle time points of the power consumption maximum value and the power consumption minimum value respectively includes:
Averaging the occurrence intervals of each adjacent power consumption maximum value in each group of the strong periodic data to obtain a first average period length of the power consumption maximum value; obtaining a periodic time point corresponding to the electricity consumption maximum value in the historical period according to the first average period length;
Averaging the occurrence intervals of each adjacent power consumption smaller value in each group of the strong periodic data to obtain a second average period length of the power consumption smaller value; and obtaining a cycle time point corresponding to the electricity consumption partial small value in the historical period according to the second average cycle length.
Further, the exceeding degree obtaining method includes:
calculating the difference value between the electricity consumption data and the average value of the electricity consumption larger value as the larger value difference degree; calculating the reciprocal of the extreme difference in the electricity consumption maximum value as the contrast of the maximum value;
Calculating the difference value between the average value of the electricity consumption smaller value and the electricity consumption data as the smaller value difference degree; calculating the reciprocal of the extremely poor in the electricity consumption smaller value as the smaller value contrast;
obtaining the exceeding degree of the electricity consumption data according to the maximum value difference degree, the maximum value contrast ratio, the minimum value difference degree and the minimum value contrast ratio; and the exceeding degree of the electricity consumption data, the maximum value difference degree, the maximum value contrast degree, the minimum value difference degree and the minimum value contrast degree are in positive correlation.
Further, obtaining the exceeding degree of the electricity consumption data according to the maximum value difference degree, the maximum value contrast ratio, the minimum value difference degree and the minimum value contrast ratio comprises the following steps:
Obtaining the exceeding degree according to a exceeding degree calculating formula, wherein the exceeding degree calculating formula is as follows:
; in the/> Representing the degree of excess of each data point in the strongly periodic data,/>Indicating the degree of exceeding of the offset value; /(I)Indicating the degree of excess of the minor value; /(I)Representing the/>, in strongly periodic dataData values for data points; /(I)Representing the number of the electricity consumption larger value; /(I)Represents the/>The power consumption is larger; /(I)Representing the maximum value of the electricity consumption partial values; /(I)Representing the minimum value of the power consumption partial values; /(I)Representing the quantity of electricity consumption with smaller value; /(I)Represents the/>The electricity consumption is smaller; /(I)Representing the maximum value of the electricity consumption smaller values; /(I)Representing the minimum value of the power consumption partial values; /(I)The parenthesis indicates Ai Fosen, and the parenthesis value is 1 if the condition in the parenthesis is satisfied, and 0 if the condition in the parenthesis is not satisfied.
Further, the method for acquiring the first weight coefficient of the weak periodic data includes:
acquiring a first weight coefficient according to a first weight coefficient calculation formula, wherein the first weight coefficient calculation formula is as follows:
; in the/> Representing the/>, within a moving window, in weak periodic dataA first weight coefficient for a data point; /(I)Representing the/>, within a moving windowData values for data points; /(I)A data average value representing the power consumption data in the moving window; /(I)Representing the number of data points contained within the moving window; /(I)Represents the/>Time sequence positions of data points in the moving window; /(I)Representing the/>, within a moving windowDistance between the data point and the current time sequence position; Representing the/>, within a moving window Degree of deviation of data points.
Further, the second weight coefficient obtaining method includes:
Obtaining a second weight coefficient according to a second weight coefficient calculation formula, wherein the second weight coefficient calculation formula is as follows:
; in the/> Representing the/>, within a moving window, in strongly periodic dataA second weight coefficient for the data point; /(I)Representing the/>, within a moving windowData values for data points; /(I)A data average value representing the power consumption data in the moving window; /(I)Representing the number of data points contained within the moving window; /(I)Represents the/>Time sequence positions of data points in the moving window; /(I)Representing the sum of the moving window/>The time sequence position of the period time point with the nearest data point; /(I)Representing the/>, within a moving windowDistance between the data point and the current time sequence position; /(I)Representing the/>, within a moving windowDegree of deviation of data points; Representing the/>, within a moving window The time sequence position of a data point and the position difference of the time period point of the latest period time point; /(I)Representing the/>, within strongly periodic dataThe degree of overrun of the data points; /(I)The parenthesis indicates Ai Fosen, and the parenthesis value is 1 if the condition in the parenthesis is satisfied, and 0 if the condition in the parenthesis is not satisfied.
Further, the preset history period is set to 3 months.
The invention has the following beneficial effects:
according to the invention, firstly, each group of electricity consumption data in a preset historical period of each electricity consumption enterprise is collected, and as the periodicity of each group of electricity consumption data is analyzed later, the electricity consumption maximum value and the electricity consumption minimum value in each group of electricity consumption data are screened out to represent the integral periodicity and facilitate the subsequent calculation of the exceeding degree of data points; then according to the respective discrete degrees of the power consumption maximum value and the power consumption minimum value, periodic parameters of power consumption data are obtained, the periodic intensity of each group of power consumption data is reflected, and then the periodic parameters are utilized to divide the power consumption data of all power consumption enterprises into strong periodic data and weak periodic data; for weak periodic data, according to the deviation degree and time sequence position of each data point in the moving window, obtaining a first weight coefficient of each data point in the moving window in the weak periodic data, and carrying out predictive analysis on the weak periodic data according to the first weight coefficient; for strong periodic data, firstly, cycle time points of a power consumption maximum value and a power consumption minimum value are obtained, and the cycle time points can reflect which time points the power consumption maximum value and the power consumption minimum value can appear; calculating the exceeding degree of each data point on the power consumption maximum value and the power consumption minimum value, wherein the exceeding degree reflects whether the abnormal degree of the data point exceeds the allowable range or not; correcting the first weight coefficient by using the time sequence position of each data point in the exceeding degree and the moving window and the position difference of the latest periodical time point, properly reducing the weight coefficient of the data point with abnormal performance, obtaining the second weight coefficient of each data point in the moving window, and carrying out predictive analysis on the strong periodical data by using the second weight coefficient. The invention considers the integral characteristics of the data, and if abnormal data occurs in the power consumption data, the weight coefficient of each data point in the moving window can be properly adjusted, so that the prediction result is more accurate.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions and advantages of the prior art, the following description will briefly explain the drawings used in the embodiments or the description of the prior art, and it is obvious that the drawings in the following description are only some embodiments of the invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for an intelligent analysis system for information for a gridding service terminal according to an embodiment of the present invention.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following detailed description refers to the specific implementation, structure, characteristics and effects of an intelligent information analysis system for a gridding service terminal according to the invention by combining the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The following specifically describes a specific scheme of the information intelligent analysis system for the gridding service terminal provided by the invention with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of a method for an intelligent information analysis system for a gridding service terminal according to an embodiment of the present invention is shown, where the method includes:
Step S1: and obtaining data of each group of electricity consumption in a preset historical period of each electricity utilization enterprise.
The embodiment of the invention mainly provides an implementation step of an intelligent analysis system for information of a gridding service terminal, and in an actual application scene, the gridding service terminal can be used for predictive analysis of electricity consumption data of an electricity enterprise. Aiming at the problem of predicting the electricity consumption data of each electricity consumption enterprise, firstly, the electricity consumption data of each electricity consumption enterprise in a preset historical period is obtained. Data acquisition functions are usually provided in modern smart meters, and can record and store electricity consumption data.
Preferably, in one embodiment of the present invention, the power consumption data may be transmitted to the management terminal through a communication interface carried on the smart meter for storage, and the preset history period is set to 3 months, and may be set by an implementation personnel, which is not limited herein.
So far, the electricity consumption data of each electricity consumption enterprise in a preset historical period are obtained.
Step S2: screening each group of electricity consumption data into two types according to the data value, and obtaining an electricity consumption larger value and an electricity consumption smaller value in each group of electricity consumption data; according to the respective discrete degrees of the power consumption maximum value and the power consumption minimum value, periodic parameters of each group of power consumption data are obtained; and dividing the power consumption data into strong periodic data and weak periodic data according to the periodic parameters.
Preferably, in one embodiment of the present invention, it is known that, because the electricity consumption data is time-series data and each electricity consumption data value in the electricity consumption data is independent, the electricity consumption data of each electricity consumption enterprise can satisfy the independence assumption of gaussian distribution. Therefore, in the embodiment of the invention, the power consumption maximum value and the power consumption minimum value of each group of power consumption data are obtained according to Gaussian distribution. The method for specifically acquiring the electricity consumption maximum value and the electricity consumption minimum value in each group of electricity consumption data comprises the following steps:
Carrying out Gaussian distribution model fitting on each group of electricity consumption data, and obtaining the mean value and standard deviation of each group of electricity consumption data according to the Gaussian distribution model; taking all data points larger than the mean value plus standard deviation as power consumption bias values of each group of power consumption data; all data points smaller than the mean value minus the standard deviation are taken as power consumption minor values of each group of power consumption data.
The gaussian distribution is a technical means well known to those skilled in the art, and only one step of obtaining the electricity consumption value and the electricity consumption value is briefly described below: in one embodiment of the invention, a data average value of the electricity consumption data of each electricity consumption enterprise is calculatedAnd standard deviation/>By means of/>, in Gaussian distributionPrinciple, will be greater than/>Is used as the power consumption maximum value point, and is smaller than/>As the power consumption amount small value point. When the periodicity of the electricity consumption data of each group is analyzed later, the periodicity of the electricity consumption maximum value and the electricity consumption minimum value can reflect the overall periodicity of the electricity consumption data of each group, so that the obtained electricity consumption maximum value and the electricity consumption minimum value are calculated later. It should be noted that, in other embodiments of the present invention, the power consumption value and the power consumption value may be automatically selected according to the specific implementation scenario, for example, the power consumption value may be greater thanIs used as the power consumption maximum value point, and is smaller than/>The data points of (2) are not limited herein as the power consumption amount small value points.
After obtaining the power consumption bias value and the power consumption bias value of each group of power consumption data, the data format needs to be specified to execute the subsequent steps of the computer program. In describing only one embodiment of the invention, use is made ofThe format represents daily electricity consumption data, and specific examples include: use/>Date,/>Representing the power usage on the day of the day:。
thus, the electricity consumption maximum value and the electricity consumption minimum value of each group of electricity consumption data are obtained.
The electricity consumption data of each group of electricity consumption enterprises may show obvious periodicity, in the electricity consumption data with obvious periodicity, the electricity consumption maximum value or the electricity consumption minimum value can appear at fixed time intervals, and because of the difference between the variation amplitude and the appearance time interval of the electricity consumption maximum value and the electricity consumption minimum value, the periodicity of the electricity consumption data has great relevance, and the difference between the variation amplitude and the appearance time interval of the electricity consumption maximum value and the electricity consumption minimum value reflects the respective discrete degree of the electricity consumption maximum value and the electricity consumption minimum value. Therefore, in the embodiment of the invention, the periodic parameters of each group of electricity consumption data are obtained according to the respective discrete degrees of the electricity consumption maximum value and the electricity consumption minimum value.
Preferably, in one embodiment of the present invention, the method for acquiring periodic parameters of power consumption data includes:
respectively calculating the integral discrete degree of all the electricity consumption larger values and the electricity consumption smaller values; the overall discrete degree is the sum of the discrete degree of the data value and the discrete degree of the occurrence time interval; and normalizing and adding the whole discrete degrees of all the power consumption larger values and the power consumption smaller values respectively to obtain the periodicity parameter. In one embodiment of the present invention, the periodic parameter calculation formula is as follows:
In the method, in the process of the invention, A periodic parameter representing each set of electricity usage data; /(I)Representing standard deviation of the occurrence time interval of the power consumption maximum value in each group of power consumption data; /(I)Standard deviation of power consumption maximum value in each group of power consumption data is represented; /(I)Representing standard deviation of the occurrence time interval of the power consumption partial value in each group of power consumption data; /(I)Representing standard deviation of power consumption smaller value in each group of power consumption data; /(I)An exponential function based on a natural constant is represented.
In one embodiment of the invention, the data discrete degree represents the data standard deviation of each of the power consumption larger value and the power consumption smaller value; the degree of discrete appearance time interval represents the standard deviation of the appearance interval of each of the power consumption larger value and the power consumption smaller value.
It should be noted that, in other embodiments of the present invention, other function mapping, maximum and minimum normalization and other basic mathematical operations may be used to implement normalization, and the specific method is a technical means well known to those skilled in the art, and is not described and limited herein.
In the calculation formula of the periodic parameter,、/>The data discrete degrees of the power consumption larger value and the power consumption smaller value are respectively represented; /(I)、/>The discrete degree of the occurrence time interval of the power consumption larger value and the power consumption smaller value is respectively represented; the data discrete degree of the electricity consumption maximum value and the occurrence time interval discrete degree are added and normalized to obtain the integral discrete degree of the electricity consumption maximum value, the influence degree of the electricity consumption maximum value on periodicity is reflected, the integral discrete degree of the electricity consumption minimum value can be obtained in the same way, and the influence degree of the electricity consumption minimum value on periodicity is reflected; and periodically adding the electricity consumption maximum value and the electricity consumption minimum value to obtain the periodic parameter of the electricity consumption data.
The periodicity of the power consumption data presentation of different power consumption enterprises is different, so that in order to facilitate the subsequent further analysis of the power consumption data with obvious periodicity, the acquired power consumption data of each group needs to be classified according to the periodicity parameters. Therefore, in the embodiment of the invention, the power consumption data is divided into strong periodic data and weak periodic data according to the periodic parameters.
Preferably, in one embodiment of the present invention, dividing the power consumption data into strong periodic data and weak periodic data according to the periodic parameter includes:
The method comprises the steps of presetting a first threshold value, taking power consumption data with the periodicity parameter larger than the first threshold value as strong periodicity data, and taking power consumption data with the periodicity parameter smaller than the first threshold value as weak periodicity data.
In one embodiment of the present invention, the preset first threshold is set to 5, that is: and traversing each group of electricity consumption data, wherein when the periodicity parameter of the electricity consumption data is smaller than 5, the electricity consumption data is weak periodicity data, and otherwise, the electricity consumption data is strong periodicity data.
So far, all the strong periodic data and the weak periodic data are obtained.
Step S3: in the calculation process of a moving average method, according to the deviation degree and time sequence position of each data point in a moving window in weak periodic data, a first weight coefficient of each data point in the moving window in the weak periodic data is obtained; in the strong periodic data, acquiring the respective periodic time points of the power consumption larger value and the power consumption smaller value; obtaining the exceeding degree of each data point in the strong periodic data on the power consumption larger value and the power consumption smaller value; and correcting the first weight coefficient by utilizing the time sequence position of each data point in the moving window and the position difference of the nearest periodical time point and the exceeding degree to obtain a second weight coefficient of each data point in the moving window in the strong periodical data.
For weak periodic data, the time sequence positions of the data points in a moving window always need to be considered when a moving average method is used for prediction, the data points which are closer to the current time sequence are given a larger weight value, and the data points which are farther from the current time sequence are given a smaller weight value; and if an abnormally large or small data point occurs in the moving window, that is, the data point deviates very greatly, the weight coefficient of the data point also needs to be reduced. Therefore, in the embodiment of the invention, the first weight coefficient of each data point in the moving window in the weak periodic data is obtained according to the deviation degree and the time sequence position of each data point in the moving window in the weak periodic data.
Preferably, in one embodiment of the present invention, the first weight coefficient obtaining method includes:
Acquiring a first weight coefficient according to a first weight coefficient calculation formula, wherein the first weight coefficient calculation formula is as follows:
In the method, in the process of the invention, Representing the/>, within a moving window, in weak periodic dataA first weight coefficient for a data point; /(I)Representing the/>, within a moving windowData values for data points; /(I)A data average value representing the power consumption data in the moving window; /(I)Representing the number of data points contained within the moving window; /(I)Represents the/>Time sequence positions of data points in the moving window; Representing the/>, within a moving window Distance between the data point and the current time sequence position; /(I)Representing the/>, within a moving windowDegree of deviation of data points.
In the first weight coefficient calculation formula,The larger the expression of the/>The more prominent the data point is relative to the data points within the moving window, the/>The greater the degree of deviation of the data points, the degree of deviation is at this point/>The smaller the (th) >The smaller the first weight coefficient of the data point; and/>The larger the move window is the/>The farther the time sequence position of the data point is from the current time sequence, the farther the data point is from the current time sequence in the moving window should be given a smaller weight coefficient in the moving average method, namely the first/>The smaller the first weight coefficient of the data point.
When the moving average method is performed on the strong periodic data, the periodicity of all the power consumption maximum values and the power consumption minimum values in the moving window needs to be analyzed, so that the respective periodic time points of the power consumption maximum values and the power consumption minimum values in the strong periodic data need to be obtained first. Therefore, in the embodiment of the invention, the cycle time points of the power consumption larger value and the power consumption smaller value are obtained.
Preferably, in one embodiment of the present invention, the method for acquiring the cycle time points of the power consumption maximum value and the power consumption minimum value includes:
Averaging the occurrence intervals of the power consumption maximization values of each adjacent power consumption in each group of strong periodic data to obtain a first average period length of the power consumption maximization values; obtaining a periodic time point corresponding to the electricity consumption maximum value in the historical period according to the first average period length; averaging the occurrence intervals of the small values of each adjacent power consumption in each group of strong periodic data to obtain a second average period length of the small values of the power consumption; and obtaining a cycle time point corresponding to the electricity consumption partial small value in the historical period according to the second average cycle length.
The time points of the cycle can be known to be the time points at which the power consumption maximum value and the power consumption minimum value occur, and if an abnormally large data point occurs at the time point of the cycle at which the power consumption maximum value is located or an abnormally small data point occurs at the time point of the cycle at which the power consumption minimum value is located, the data points which are abnormal in performance cannot be used as references in the prediction of the power consumption, and the weight coefficients of the data points which are abnormal in performance should be reduced. At this time, it is determined whether the data point should be reduced by the weight coefficient. Therefore, in the embodiment of the invention, the exceeding degree of each data point in the strong periodic data to the power consumption larger value and the power consumption smaller value is obtained.
Preferably, in one embodiment of the present invention, the exceeding degree obtaining method includes:
Calculating a difference value between the electricity consumption data and the average value of the electricity consumption maximum value as a maximum value difference degree; calculating the reciprocal of the extreme difference in the electricity consumption maximum value as the contrast of the maximum value; calculating a difference value between the average value of the electricity consumption smaller value and the electricity consumption data as a smaller value difference degree; calculating the reciprocal of the extremely poor in the electricity consumption extremely small value as the contrast of the extremely small value; obtaining the exceeding degree of the power consumption data according to the difference degree of the maximum value, the contrast ratio of the maximum value, the difference degree of the minimum value and the contrast ratio of the minimum value; the exceeding degree of the power consumption data and the difference degree of the maximum value, the contrast ratio of the maximum value, the difference degree of the minimum value and the contrast ratio of the minimum value all have positive correlation.
Preferably, in one embodiment of the present invention, obtaining the exceeding degree of the electricity consumption data according to the maximum value difference degree, the maximum value contrast ratio, the minimum value difference degree and the minimum value contrast ratio includes:
obtaining the exceeding degree according to a exceeding degree calculating formula, wherein the exceeding degree calculating formula is as follows:
In the method, in the process of the invention, Representing the degree of excess of each data point in the strongly periodic data,/>Indicating the degree of exceeding of the offset value; /(I)Indicating the degree of excess of the minor value; /(I)Representing the/>, in strongly periodic dataData values for data points; /(I)Representing the number of the electricity consumption larger value; /(I)Represents the/>The power consumption is larger; /(I)Representing the maximum value of the electricity consumption partial values; /(I)Representing the minimum value of the power consumption partial values; /(I)Representing the quantity of electricity consumption with smaller value; /(I)Represents the/>The electricity consumption is smaller; Representing the maximum value of the electricity consumption smaller values; /(I) Representing the minimum value of the power consumption partial values; /(I)The parenthesis indicates Ai Fosen, and the parenthesis value is 1 if the condition in the parenthesis is satisfied, and 0 if the condition in the parenthesis is not satisfied.
In the excess degree calculation formula, when the abnormally large data value is larger than the average value of the power consumption amount deviation value, the abnormally large data value is considered to be out of the abnormal allowable range, at this time, the value in the Ai Fosen brackets is 1,Representing the/>, in strongly periodic dataThe degree of prominence of data points to larger values of electricity consumption,/>Negative correlation mapping is performed on the overall contrast representing the electricity consumption value, and the smaller the overall contrast is, namely/>The smaller the overall data trend of the larger value of the electricity consumption is, the more stable the highlighting degree should be enhanced, so the/>, in the strong periodic data isThe greater the degree of exceeding of the data point to the power consumption maximum value; similarly,/>The larger and/>The smaller the/>, the more periodic the dataThe larger the degree of the data point exceeding the smaller value of the smaller electricity consumption; in addition, the smaller power consumption value may occur at the period time point of the larger power consumption value or the larger power consumption value occurs at the period time point of the smaller power consumption value, and the data points are considered not to exceed the allowable range of abnormality, and the data points are brought into the formula of the exceeding degree and do not meet the condition in Ai Fosen brackets, so that the exceeding degree of the data points is 0.
Whether for strong periodic data or weak periodic data, the time sequence position of the data point in the moving window and the deviation degree of the data point always need to be considered when the prediction is carried out by using a moving average method; however, for strong periodic data, the adjustment of the weight also needs to consider the period of the data points in the moving window, the distance between the period time point where the power consumption is larger and the period time point where the power consumption is smaller and the period time point where each data point is located is used for correcting the weight coefficient of each data point, and if the data point exceeds the abnormal allowable range, the weight coefficient also needs to be further corrected by using the exceeding degree of the data point. Therefore, in the embodiment of the invention, the second weight coefficient of each data point in the strong periodic data moving window is obtained.
Preferably, in one embodiment of the present invention, the second weight coefficient obtaining method includes:
Obtaining a second weight coefficient according to a second weight coefficient calculation formula, wherein the second weight coefficient calculation formula is as follows:
In the method, in the process of the invention, Representing the/>, within a moving window, in strongly periodic dataWeighting coefficients for the data points; /(I)Representing the/>, within a moving windowData values for data points; /(I)A data average value representing the power consumption data in the moving window; /(I)Representing the number of data points contained within the moving window; /(I)Represents the/>Time sequence positions of data points in the moving window; /(I)Representing the sum of the moving window/>The time sequence position of the period time point with the nearest data point; /(I)Representing the/>, within a moving windowDistance between the data point and the current time sequence position; /(I)Representing the/>, within a moving windowDegree of deviation of data points; /(I)Representing the/>, within a moving windowThe time sequence position of the data point and the position difference of the time period point of the latest period time point; /(I)Representing the/>, within strongly periodic dataThe degree of overrun of the data points; /(I)The parenthesis indicates Ai Fosen, and the parenthesis value is 1 if the condition in the parenthesis is satisfied, and the parenthesis value is 0 deviation degree if the condition in the parenthesis is not satisfied.
In the weight coefficient calculation formula,Representing the/>, within a moving windowThe difference between the time sequence position of the data point and the position of the latest cycle time point should be considered with respect to the power consumption larger value and the power consumption smaller value which do not appear at the corresponding cycle time pointsCorrecting the position difference between the period time points with the nearest data point positions, wherein the greater the position difference is, the smaller the degree of correcting the weight coefficient is, and the first/>, in the moving window isThe smaller the second weight coefficient increase amplitude of the data point; in combination with the degree of overrun, the larger the degree of overrun, the smaller the magnitude of increase of the second weight coefficient should be.
Thus, a first weight coefficient of each data point in the moving window in the weak periodic data and a second weight coefficient of each data point in the moving window in the strong periodic data are obtained.
Step S4: and according to the first weight coefficient and the second weight coefficient, carrying out predictive analysis on the electricity consumption data by using a moving average method.
When predicting future power usage using a moving average method, it is necessary to obtain the weight coefficients of all the data points in the moving window. And S3, dynamically adjusting weight coefficients of all data points in the moving window, substituting the adjusted weight coefficients into a moving average method calculation process, and obtaining a predicted value of power consumption data of a day in the future at the current time point. The power consumption enterprises can reasonably arrange the power consumption scheduling of the enterprises according to the predicted value of the power consumption data obtained by the moving average method as a reference.
It should be noted that the moving average method is a technical means well known to those skilled in the art, and is not described herein, and in one embodiment of the present invention, the number of data points included in the moving window is set to be 5.
In summary, the present invention collects each group of electricity consumption data in a preset history period of each electricity consumption enterprise; then according to the discrete degree of each of the power consumption maximum value and the power consumption minimum value, periodic parameters of power consumption data are obtained, and then all groups of power consumption data are divided into strong periodic data and weak periodic data by utilizing the periodic parameters; for weak periodic data, according to the deviation degree and time sequence position of each data point in the moving window, obtaining a first weight coefficient of each data point in the moving window in the weak periodic data, and carrying out predictive analysis on the weak periodic data according to the first weight coefficient; for strong periodic data, firstly, a periodic time point of a power consumption maximum value and a power consumption minimum value is obtained; calculating the exceeding degree of each data point on the power consumption maximum value and the power consumption minimum value; correcting the first weight coefficient by using the time sequence position of each data point in the exceeding degree and the moving window and the position difference of the latest periodical time point, properly reducing the weight coefficient of the data point with abnormal expression, obtaining the second weight coefficient of each data point in the moving window, and carrying out predictive analysis on the strong periodical data by using the second weight coefficient. The invention considers the integral characteristics of the data, and if abnormal data occurs in the power consumption data, the weight coefficient of each data point in the moving window can be properly adjusted, so that the prediction result is more accurate.
It should be noted that: the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from other embodiments.
Claims (2)
1. An intelligent information analysis system for a gridding service terminal, comprising a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor executes the computer program to realize the following steps:
acquiring data of each group of electricity consumption in a preset historical period of each electricity consumption enterprise;
screening each group of electricity consumption data into two types according to the data value, and obtaining an electricity consumption maximum value and an electricity consumption minimum value in each group of electricity consumption data; according to the discrete degree of each of the electricity consumption maximum value and the electricity consumption minimum value, periodic parameters of each group of electricity consumption data are obtained; dividing the electricity consumption data into strong periodic data and weak periodic data according to the periodic parameters;
In the calculation process of a moving average method, according to the deviation degree and time sequence position of each data point in a moving window in weak periodic data, a first weight coefficient of each data point in the moving window in the weak periodic data is obtained; acquiring the cycle time points of the power consumption maximum value and the power consumption minimum value in the strong periodic data; obtaining the exceeding degree of each data point in the strong periodic data on the power consumption larger value and the power consumption smaller value; correcting the first weight coefficient by utilizing the time sequence position of each data point in the moving window and the position difference of the nearest periodic time point and the exceeding degree to obtain a second weight coefficient of each data point in the moving window in strong periodic data;
According to the first weight coefficient and the second weight coefficient, carrying out predictive analysis on the electricity consumption data by using a moving average method;
the method for acquiring the electricity consumption maximum value and the electricity consumption minimum value in each group of electricity consumption data comprises the following steps:
Carrying out Gaussian distribution model fitting on each group of electricity consumption data, and obtaining the mean value and standard deviation of each group of electricity consumption data according to the Gaussian distribution model;
taking all data points larger than the mean value plus standard deviation as the electricity consumption maximum value of each group of electricity consumption data; taking all data points smaller than the mean value minus the standard deviation as the electricity consumption partial value of each group of electricity consumption data;
The periodic parameter acquisition method of the electricity consumption data comprises the following steps:
respectively calculating the overall discrete degree of all the electricity consumption maximum values and the electricity consumption minimum values; the integral discrete degree is the sum of the discrete degree of the data value and the discrete degree of the occurrence time interval;
Normalizing and adding the overall discrete degrees of all the power consumption maximum values and the power consumption minimum values respectively to obtain the periodic parameters;
Dividing the electricity consumption data into strong periodic data and weak periodic data according to the periodic parameters, wherein the method comprises the following steps:
presetting a first threshold, taking the power consumption data with the periodic parameter larger than the first threshold as strong periodic data, and taking the power consumption data with the periodic parameter smaller than the first threshold as weak periodic data;
the method for acquiring the cycle time points of the power consumption maximum value and the power consumption minimum value comprises the following steps:
Averaging the occurrence intervals of each adjacent power consumption maximum value in each group of the strong periodic data to obtain a first average period length of the power consumption maximum value; obtaining a periodic time point corresponding to the electricity consumption maximum value in the historical period according to the first average period length;
Averaging the occurrence intervals of each adjacent power consumption smaller value in each group of the strong periodic data to obtain a second average period length of the power consumption smaller value; obtaining a periodic time point corresponding to the electricity consumption smaller value in the historical period according to the second average periodic length;
The exceeding degree obtaining method comprises the following steps:
calculating the difference value between the electricity consumption data and the average value of the electricity consumption larger value as the larger value difference degree; calculating the reciprocal of the extreme difference in the electricity consumption maximum value as the contrast of the maximum value;
Calculating the difference value between the average value of the electricity consumption smaller value and the electricity consumption data as the smaller value difference degree; calculating the reciprocal of the extremely poor in the electricity consumption smaller value as the smaller value contrast;
Obtaining the exceeding degree of the electricity consumption data according to the maximum value difference degree, the maximum value contrast ratio, the minimum value difference degree and the minimum value contrast ratio; the exceeding degree of the electricity consumption data and the maximum value difference degree, the maximum value contrast degree, the minimum value difference degree and the minimum value contrast degree all have positive correlation relations;
Obtaining the exceeding degree of the electricity consumption data according to the maximum value difference degree, the maximum value contrast degree, the minimum value difference degree and the minimum value contrast degree, wherein the method comprises the following steps:
Obtaining the exceeding degree according to a exceeding degree calculating formula, wherein the exceeding degree calculating formula is as follows:
; in the method, in the process of the invention, Representing the degree of excess of each data point in the strongly periodic data,/>Indicating the degree of exceeding of the offset value; /(I)Indicating the degree of excess of the minor value; /(I)Representing the/>, in strongly periodic dataData values for data points; /(I)Representing the number of the electricity consumption larger value; represents the/> The power consumption is larger; /(I)Representing the maximum value of the electricity consumption partial values; /(I)Representing the minimum value of the power consumption partial values; /(I)Representing the quantity of electricity consumption with smaller value; /(I)Represents the/>The electricity consumption is smaller; /(I)Representing the maximum value of the electricity consumption smaller values; /(I)Representing the minimum value of the power consumption partial values; /(I)A Ai Fosen bracket, wherein the bracket value is 1 if the condition in the bracket is satisfied, and 0 if the condition in the bracket is not satisfied;
The first weight coefficient acquisition method of the weak periodic data comprises the following steps:
acquiring a first weight coefficient according to a first weight coefficient calculation formula, wherein the first weight coefficient calculation formula is as follows:
; in the/> Representing the/>, within a moving window, in weak periodic dataA first weight coefficient for a data point; /(I)Representing the/>, within a moving windowData values for data points; /(I)A data average value representing the power consumption data in the moving window; /(I)Representing the number of data points contained within the moving window; /(I)Represents the/>Time sequence positions of data points in the moving window; /(I)Representing the/>, within a moving windowDistance between the data point and the current time sequence position; Representing the/>, within a moving window Degree of deviation of data points;
The second weight coefficient acquisition method comprises the following steps:
Obtaining a second weight coefficient according to a second weight coefficient calculation formula, wherein the second weight coefficient calculation formula is as follows:
; in the/> Representing the/>, within a moving window, in strongly periodic dataA second weight coefficient for the data point; /(I)Representing the/>, within a moving windowData values for data points; /(I)A data average value representing the power consumption data in the moving window; /(I)Representing the number of data points contained within the moving window; represents the/> Time sequence positions of data points in the moving window; /(I)Representing the sum of the moving window/>The time sequence position of the period time point with the nearest data point; /(I)Representing the/>, within a moving windowDistance between the data point and the current time sequence position; /(I)Representing the/>, within a moving windowDegree of deviation of data points; Representing the/>, within a moving window The time sequence position of a data point and the position difference of the time period point of the latest period time point; /(I)Representing the/>, within strongly periodic dataThe degree of overrun of the data points; /(I)The parenthesis indicates Ai Fosen, and the parenthesis value is 1 if the condition in the parenthesis is satisfied, and 0 if the condition in the parenthesis is not satisfied.
2. The intelligent analysis system for information for a gridding service terminal according to claim 1, wherein the preset history period is set to 3 months.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410205940.8A CN117786370B (en) | 2024-02-26 | 2024-02-26 | Information intelligent analysis system for gridding service terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410205940.8A CN117786370B (en) | 2024-02-26 | 2024-02-26 | Information intelligent analysis system for gridding service terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117786370A CN117786370A (en) | 2024-03-29 |
| CN117786370B true CN117786370B (en) | 2024-04-19 |
Family
ID=90393000
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410205940.8A Active CN117786370B (en) | 2024-02-26 | 2024-02-26 | Information intelligent analysis system for gridding service terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117786370B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11487273B1 (en) * | 2021-04-30 | 2022-11-01 | Dalian University Of Technology | Distributed industrial energy operation optimization platform automatically constructing intelligent models and algorithms |
| CN116418882A (en) * | 2023-06-09 | 2023-07-11 | 北京国旺盛源智能终端科技有限公司 | Memory data compression method based on HPLC dual-mode carrier communication |
| CN116541666A (en) * | 2023-05-18 | 2023-08-04 | 中国建筑科学研究院有限公司 | Low-carbon park carbon tracking method based on influence factor tracing |
| CN117035154A (en) * | 2023-06-25 | 2023-11-10 | 武汉大学 | Multi-element load prediction method and device for comprehensive energy system |
-
2024
- 2024-02-26 CN CN202410205940.8A patent/CN117786370B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11487273B1 (en) * | 2021-04-30 | 2022-11-01 | Dalian University Of Technology | Distributed industrial energy operation optimization platform automatically constructing intelligent models and algorithms |
| CN116541666A (en) * | 2023-05-18 | 2023-08-04 | 中国建筑科学研究院有限公司 | Low-carbon park carbon tracking method based on influence factor tracing |
| CN116418882A (en) * | 2023-06-09 | 2023-07-11 | 北京国旺盛源智能终端科技有限公司 | Memory data compression method based on HPLC dual-mode carrier communication |
| CN117035154A (en) * | 2023-06-25 | 2023-11-10 | 武汉大学 | Multi-element load prediction method and device for comprehensive energy system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117786370A (en) | 2024-03-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11876374B2 (en) | System and method for optimal control of energy storage system | |
| CN116646933B (en) | Big data-based power load scheduling method and system | |
| US20040225625A1 (en) | Method and system for calculating and distributing utility costs | |
| CN113283674A (en) | Baseline load prediction correction method based on user electricity utilization characteristics | |
| WO2020053846A2 (en) | A system and method for analysis of smart meter data | |
| CN112907062A (en) | Power grid electric quantity prediction method, device, medium and terminal integrating temperature characteristics | |
| CN114336651A (en) | Power dispatching method and device based on peak clipping potential | |
| CN113191575A (en) | Power distribution network maintenance power failure mode optimization method and device, terminal and storage medium | |
| CN116050866A (en) | Combined user missing electric quantity fitting method and system based on typical load curve | |
| CN117786370B (en) | Information intelligent analysis system for gridding service terminal | |
| CN111708987A (en) | Method for predicting load of multiple parallel transformers of transformer substation | |
| Guan et al. | Advanced load forecast with hierarchical forecasting capability | |
| CN113826294A (en) | Power management method, apparatus, computing device, medium, and product | |
| CN117650628B (en) | Energy efficiency management system based on self-adaptive preconditioning scene | |
| CN117421690B (en) | Intelligent park energy consumption real-time monitoring method and system | |
| CN110598899B (en) | Local electric energy meter Shen Jiaofu business planning method and system based on data prediction | |
| CN117593171B (en) | Image acquisition, storage and processing method based on FPGA | |
| CN117895509B (en) | Multi-energy coupling type power distribution network scheduling method and system | |
| Isaac et al. | Qbsd: Quartile-Based Seasonality Decomposition for Cost-Effective Time Series Forecasting | |
| CN115528687B (en) | Power system flexible response capability optimization method under limited cost constraint | |
| CN110297145B (en) | Voltage sag detection method based on multi-user electric energy data deep analysis | |
| US20230155387A1 (en) | Systems and methods for dynamic tuning of controller in power grids | |
| Leithon et al. | Optimization Strategy for Energy Allocation through Cooperative Storage Management | |
| CN117439042A (en) | Multiple regression load prediction method considering temperature accumulation effect | |
| CN117856234A (en) | Power load prediction method and system based on demand side response |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |