CN113887831A - Novel power load prediction influence factor correlation analysis method - Google Patents
Novel power load prediction influence factor correlation analysis method Download PDFInfo
- Publication number
- CN113887831A CN113887831A CN202111255136.3A CN202111255136A CN113887831A CN 113887831 A CN113887831 A CN 113887831A CN 202111255136 A CN202111255136 A CN 202111255136A CN 113887831 A CN113887831 A CN 113887831A
- Authority
- CN
- China
- Prior art keywords
- interval
- load
- rainfall
- temperature
- humidity
- 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.)
- Pending
Links
- 238000010219 correlation analysis Methods 0.000 title claims abstract description 29
- 238000000034 method Methods 0.000 title claims description 35
- 108010014173 Factor X Proteins 0.000 claims description 16
- 238000004458 analytical method Methods 0.000 claims description 7
- 238000005192 partition Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 5
- 230000005611 electricity Effects 0.000 claims description 4
- 238000013480 data collection Methods 0.000 claims description 3
- 238000004422 calculation algorithm Methods 0.000 abstract description 12
- 238000011161 development Methods 0.000 description 3
- 238000010801 machine learning Methods 0.000 description 3
- 238000012098 association analyses Methods 0.000 description 2
- 238000012731 temporal analysis Methods 0.000 description 2
- 238000000700 time series analysis Methods 0.000 description 2
- 230000003442 weekly effect Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 230000001537 neural effect Effects 0.000 description 1
- 238000010248 power generation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Theoretical Computer Science (AREA)
- Strategic Management (AREA)
- Data Mining & Analysis (AREA)
- Human Resources & Organizations (AREA)
- Mathematical Physics (AREA)
- Mathematical Optimization (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Operations Research (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Water Supply & Treatment (AREA)
- Algebra (AREA)
- Public Health (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention provides a novel load prediction influence factor correlation analysis algorithm. Firstly, constructing an influence factor set of load prediction, wherein the set comprises 5 major elements, temperature, humidity, day type, weather and rainfall; secondly, carrying out interval division on the influence factors and the load; then, a sample set of the load and the influence factors is constructed, wherein each sample comprises 6 major elements { temperature belonging interval, humidity belonging interval, day type, weather type, rainfall belonging interval, load belonging interval }; then, based on the measured data, calculating the occurrence probability of each sample; and finally, analyzing factors influencing the load based on the sample probability. The algorithm adopted by the invention analyzes the probability of the influence factors on the load in the big data sample, and can effectively analyze whether the influence factors generate linear influence or nonlinear influence on the load. The load prediction influence factors considered by the invention are rich.
Description
Technical Field
The invention relates to a novel power load prediction influence factor correlation analysis method, and belongs to the field of power.
Background
Power load prediction is one of the important jobs in the power sector. The accurate load prediction can ensure the safety and stability of the operation of the power grid, ensure the normal production and life of the society, effectively reduce the power generation cost and improve the economic benefit and the social benefit. The time scale of short-term load prediction is from hour level to week, generally referred to as daily load prediction and weekly load prediction, and is respectively used for arranging daily scheduling plans and weekly scheduling plans, including determining unit start and stop, water, fire and electricity coordination, tie line exchange power, load economic distribution, reservoir scheduling, equipment maintenance and the like.
The current load prediction technology mainly relates to three methods: time series analysis, correlation analysis, and machine learning algorithms. The time series analysis is a quantitative regression prediction method, and on the basis of the continuity of the admitted development, the development trend is presumed by applying the statistical analysis of historical data; in addition, the randomness generated due to the influence of accidental factors is fully considered, and in order to eliminate the influence of random fluctuation, data is appropriately processed and trend prediction is carried out. The correlation analysis is used for quantifying the degree of correlation between correlation variables, uncertain and loose dependency relations exist between the correlation variables, for a certain value of the variable, a plurality of values of another variable can correspond to the variable, and the values show regular fluctuation around the average number of the values, so that the research method for predicting is the correlation analysis prediction method. The machine learning algorithm is a core method of artificial intelligence and neural computation, and rules and values behind load data can be mined to a greater extent through learning after clustering.
It is emphasized that the three techniques are not isolated from each other in load prediction, for example, before load prediction is performed by using a machine learning algorithm, a correlation analysis algorithm is used to select factors influencing the load. Common influencing factors are spatial correlation factors and temporal periodicity factors. The space factors mainly include various load units, regional planning layout, building design characteristics and the like; the time sequence factors mainly comprise various load sequences, social development level, day types, meteorological conditions and the like.
The prior art has the following disadvantages:
1. from the aspect of algorithm, the existing impact factor association analysis algorithm for load prediction is generally based on a pearson correlation coefficient, the pearson correlation coefficient is used for measuring a linear relation between variables, and when the impact factor for load prediction has a nonlinear influence on load data, the pearson correlation coefficient fails. Moreover, the pearson correlation coefficient is used for measuring the degree of correlation between two variables, and the influence of a plurality of influence factors on the load cannot be analyzed at one time.
2. In view of influence factors, in the existing load prediction influence factor set, the considered factors are single, and the influence of multi-dimensional data including temperature, humidity, weather, rainfall and day type on the load is not considered.
Disclosure of Invention
The invention aims to provide a novel power load prediction influence factor correlation analysis method and a power load prediction method based on the influence factor correlation analysis, so that no matter linear influence or nonlinear influence is generated on a load by the influence factors, the influence factors can be effectively analyzed.
The invention adopts the following technical scheme:
a novel power load prediction influence factor correlation analysis method is characterized by comprising the following steps:
step one, interval division:
acquiring measured data of temperature, humidity, day type, weather, rainfall and load, and performing interval division of the data;
step two, counting the probability distribution of the influence factors about the load:
counting the probability distribution of temperature, humidity, day type, weather and rainfall data about load data, and recording as PLi,XjWherein i represents that the load is in the ith load interval; x represents an influence factor, T, W, D, K and R in practical application respectively represent temperature, humidity, day type, weather and rainfall; j indicates that the influence factor is in the jth interval;
step three, calculating a conditional probability set of the influence factors:
on the premise that the load is in the ith interval, the probability that the influence factor X is in the jth interval is calculated, and the calculation formula is
Fourthly, calculating a correlation coefficient:
defining a coefficient of correlation K of the influencing factors to the loadXCalculated as follows:
fifthly, judging the influence factors:
when K isXWhen the load is more than or equal to 0.8, the influence factor X is judged as an effective influence factor of the load prediction and can be used for the load prediction;
when K isXIf the load factor is less than 0.8, the influence factor X is judged as an invalid influence factor of the load prediction and does not need to be considered in the load prediction.
Further, the novel power load prediction influence factor correlation analysis method of the invention also has the following characteristics:
in the first step, the partition mode is as follows:
the lowest temperature in the temperature sample data is T1Maximum temperature of T2Temperature is divided into N at equal intervalsTAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith temperature interval is defined as the ratio of the number of temperature samples belonging to the temperature interval i to all temperature samples, and is counted as PTi;
The lowest humidity in the humidity sample data is W1Maximum humidity of W2Equally dividing humidity into NWAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith humidity interval is defined as the ratio of the number of humidity samples belonging to the humidity interval i to all the humidity samples, and is counted as PWi;
The date type sample data has two intervalsThe interval 1 is the working day and is marked as D1And interval 2 is a non-working day and is marked as D2. The probability of the ith day type interval is defined as the proportion of the number of the day type samples belonging to the day type interval i to all the day type samples, and is counted as PDi;
The weather sample data has 5 sections, wherein the section 1 is sunny, the section 2 is cloudy, the section 3 is cloudy, the section 4 is rain, and the section 5 is snow, which are respectively marked as K1,K2,K3,K4,K5(ii) a The higher the sequence number of the interval is, the higher the priority is, for example, the weather of cloudy to sunny and cloudy to sunny falls into the interval 2; the probability of the ith weather interval is defined as the proportion of the number of the weather samples belonging to the weather interval i to all the weather samples, and is counted as PKi;
The minimum rainfall in the rainfall sample data is R1The maximum rainfall is R2Equally dividing the rainfall into NRAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith rainfall interval is defined as the proportion of the number of rainfall samples belonging to the rainfall interval i to all the rainfall samples, and is counted as PRi;
The lowest load in the load sample data is L1The highest load is L2Will beLoad sampleEqually spaced division into NLAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith load interval is defined as the proportion of the number of load samples belonging to the load interval i to all the load samples, and is counted as PLi。
The invention also provides a power load prediction method based on the correlation analysis of the influence factors, which is characterized by comprising the following steps:
step one, a first data acquisition module acquires the daily electricity consumption data and the daily influence factor change information,
the second data acquisition module acquires power utilization data before the day and influence factor change information before the day;
and step two, analyzing the relevance of the influence factors and the load prediction by an influence factor relevance analysis module, wherein the analysis comprises the following steps:
step 2-1, partition
Acquiring measured data of temperature, humidity, day type, weather, rainfall and load, and completing data interval division according to a method in index parameter definition.
Step 2-2, counting probability distribution of influence factors about load
Counting the probability distribution of temperature, humidity, day type, weather and rainfall data about load data, and recording as PLi,XjWherein i represents that the load is in the ith load interval; x represents an influence factor, T, W, D, K and R in practical application respectively represent temperature, humidity, day type, weather and rainfall; j indicates that the influence factor is in the jth interval. For example, the probability that the temperature is in the 1 st interval of the temperature and the load is in the 1 st interval of the load is PL1,T1。
Step 2-3, calculating a conditional probability set of the influence factors
On the premise that the load is in the ith interval, the probability that the influence factor X is in the jth interval is calculated, and the calculation formula is
Step 2-4, calculating the correlation coefficient
Defining a coefficient of correlation K of the influencing factors to the loadX. Calculated as follows.
Step 2-5, judging influence factors
When K isXWhen the load is more than or equal to 0.8, the influence factor X is judged as an effective influence factor of the load prediction and can be used for the load prediction;
when K isXIf the load is less than 0.8, the influence factor X is judged as an invalid influence factor of the load prediction, and the factors are not required to be considered when the load prediction is carried out in the step three;
and step three, according to the effective influence factors obtained in the step two, the power load prediction module predicts the power load of the remaining time of the day, or predicts the power load of the next day and the future preset time period.
The power load prediction method based on the influence factor correlation analysis of the invention also has the following characteristics:
the partition in step 2-1 is as follows:
the lowest temperature in the temperature sample data is T1Maximum temperature of T2Temperature is divided into N at equal intervalsTAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith temperature interval is defined as the ratio of the number of temperature samples belonging to the temperature interval i to all temperature samples, and is counted as PTi;
The lowest humidity in the humidity sample data is W1Maximum humidity of W2Equally dividing humidity into NWAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith humidity interval is defined as the ratio of the number of humidity samples belonging to the humidity interval i to all the humidity samples, and is counted as PWi;
The date type sample data has two intervals, wherein the interval 1 is a working day and is marked as D1And interval 2 is a non-working day and is marked as D2. The probability of the ith day type interval is defined as the proportion of the number of the day type samples belonging to the day type interval i to all the day type samples, and is counted as PDi;
The weather sample data has 5 intervals, wherein the interval 1 is sunny, the interval 2 is cloudy, the interval 3 is cloudy, and the interval 4 is cloudyRain, snow in the interval 5, respectively denoted as K1,K2,K3,K4,K5(ii) a The higher the sequence number of the interval is, the higher the priority is, for example, the weather of cloudy to sunny and cloudy to sunny falls into the interval 2; the probability of the ith weather interval is defined as the proportion of the number of the weather samples belonging to the weather interval i to all the weather samples, and is counted as PKi;
The minimum rainfall in the rainfall sample data is R1The maximum rainfall is R2Equally dividing the rainfall into NRAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith rainfall interval is defined as the proportion of the number of rainfall samples belonging to the rainfall interval i to all the rainfall samples, and is counted as PRi;
The lowest load in the load sample data is L1The highest load is L2Will beLoad sampleEqually spaced division into NLAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith load interval is defined as the proportion of the number of load samples belonging to the load interval i to all the load samples, and is counted as PLi。
Further, the method for predicting the power load based on the correlation analysis of the influence factors of the present invention has the following characteristics:
and the system also comprises an associated factor adding module for receiving other factors besides temperature, humidity, day type, weather and rainfall.
Further, the method for predicting the power load based on the correlation analysis of the influence factors of the present invention has the following characteristics:
and the newly added correlation factors in the correlation factor adding module enter the step two to analyze the correlation with the load prediction.
Further, the method for predicting the power load based on the correlation analysis of the influence factors of the present invention has the following characteristics:
and when the data before the current day collected in the second data collection module is accumulated to a preset time, the second step is carried out again, and the probability distribution of all the influencing factors about the load is counted.
The invention has the beneficial effects that: 1. the algorithm adopted by the invention analyzes the probability of the influence factors on the load in the big data sample, and can effectively analyze whether the influence factors generate linear influence or nonlinear influence on the load.
2. The method has rich considered load forecasting influence factors, and can simultaneously analyze the influence of multi-dimensional data of temperature, humidity, weather, rainfall and day type on the load. And the factor with the largest correlation with the load change is screened out, and the factor with smaller correlation is eliminated, so that the analysis workload is reduced, and the prediction accuracy is improved.
Detailed Description
The technical solution of the present invention will be further described below by way of specific embodiments.
The novel load prediction influence factor correlation analysis method provided by the embodiment. Firstly, constructing an influence factor set for load prediction, wherein the set comprises 5 influence factors: temperature, humidity, type of day, weather and rainfall; secondly, carrying out interval division on the influence factors and the load; next, a sample set of load and influence factors is constructed, each sample containing 6 major elements: temperature belonging interval, humidity belonging interval, day type, weather type, rainfall belonging interval and load belonging interval; then, based on the measured data, calculating the occurrence probability of each sample; and finally, analyzing factors influencing the load based on the sample probability.
The load prediction influence factor correlation analysis algorithm comprises influence factor indexes of temperature, humidity, day type, weather and rainfall and load data indexes, a sample set of loads and influence factors is constructed by various indexes, and data support is provided for further judging the basis of influencing load prediction.
Index type definition and index parameter definition
(1) Index type definition
Temperature, humidity, day type, weather, rainfall, load.
(2) Index parameter definition
The lowest temperature in the temperature sample data is T1Maximum temperature of T2Temperature is divided into N at equal intervalsTAn i-th interval ofIt is noted that the right side of the last interval is also a closed interval. The probability of the ith temperature interval is defined as the ratio of the number of temperature samples belonging to the temperature interval i to all temperature samples, and is counted as PTi。
The lowest humidity in the humidity sample data is W1Maximum humidity of W2Equally dividing humidity into NWAn i-th interval ofIt is noted that the right side of the last interval is also a closed interval. The probability of the ith humidity interval is defined as the ratio of the number of humidity samples belonging to the humidity interval i to all the humidity samples, and is counted as PWi。
The date type sample data has two intervals, wherein the interval 1 is a working day and is marked as D1And interval 2 is a non-working day and is marked as D2. The probability of the ith day type interval is defined as the proportion of the number of the day type samples belonging to the day type interval i to all the day type samples, and is counted as PDi。
The weather sample data has 5 sections, wherein the section 1 is sunny, the section 2 is cloudy, the section 3 is cloudy, the section 4 is rain, and the section 5 is snow, which are respectively marked as K1,K2,K3,K4,K5. The higher the sequence number of the interval is, the higher the priority is, for example, the weather of cloudy to sunny and cloudy to sunny falls into the interval 2. The probability of the ith weather interval is defined as the proportion of the number of the weather samples belonging to the weather interval i to all the weather samples, and is counted as PKi。
The minimum rainfall in the rainfall sample data is R1The maximum rainfall is R2Equally dividing the rainfall into NRAn i-th interval ofIt is noted that the right side of the last interval is also a closed interval. The probability of the ith rainfall interval is defined as the proportion of the number of rainfall samples belonging to the rainfall interval i to all the rainfall samples, and is counted as PRi。
The lowest load in the load sample data is L1The highest load is L2Will beLoad sampleEqually spaced division into NLAn i-th interval ofIt is noted that the right side of the last interval is also a closed interval. The probability of the ith load interval is defined as the proportion of the number of load samples belonging to the load interval i to all the load samples, and is counted as PLi。
(II) Algorithm flow
The steps of a novel load prediction influence factor association analysis algorithm are introduced as follows:
first step, interval division
Acquiring measured data of temperature, humidity, day type, weather, rainfall and load, and completing data interval division according to a method in index parameter definition.
Second, the probability distribution of statistical influencing factors with respect to the load
Counting the probability distribution of temperature, humidity, day type, weather and rainfall data about load data, and recording as PLi,XjWherein i represents that the load is in the ith load interval; x represents an influence factor, T, W, D, K and R in practical application respectively represent temperature, humidity, day type, weather and rainfall; j indicates that the influence factor is in the jth interval. For example, the temperature is in the 1 st section of the temperature and the load is in the 1 st section of the loadThe ratio is PL1,T1。
Thirdly, calculating a conditional probability set of the influencing factors
On the premise that the load is in the ith interval, the probability that the influence factor X is in the jth interval is calculated, and the calculation formula is
Fourthly, calculating the correlation coefficient
Defining a coefficient of correlation K of the influencing factors to the loadX. Calculated as follows.
Fifth, judgment of influence factors
When K isXAnd when the influence factor X is more than or equal to 0.8, judging the influence factor X as an effective influence factor of the load prediction, and using the effective influence factor for the load prediction.
When K isXIf the load factor is less than 0.8, the influence factor X is judged as an invalid influence factor of the load prediction and does not need to be considered in the load prediction.
The novel load prediction influence factor correlation analysis method is used as a core, and the power load prediction method based on the influence factor correlation analysis is constructed as follows:
step one, a first data acquisition module acquires the daily electricity consumption data and the daily influence factor change information,
the second data acquisition module acquires power utilization data before the day and influence factor change information before the day;
step two, the influence factor correlation analysis module analyzes the correlation between the influence factors and the load prediction,
step 2-1, partition
Step 2-2, counting probability distribution of influence factors about load
Step 2-3, calculating a conditional probability set of the influence factors
Step 2-4, calculating the correlation coefficient
Step 2-5, judging influence factors
And step three, according to the effective influence factors obtained in the step two, the power load prediction module predicts the power load of the remaining time of the day, or predicts the power load of the next day and the future preset time period. The power load prediction method can be realized by adopting the prediction methods in the prior art, and the methods predict based on the related factors obtained by the related analysis of the influence factors in the first step and the second step, so that the prediction accuracy can be effectively improved.
In some embodiments, after the data collected in the second data collection module before the current day is accumulated for a predetermined time, the second step is performed again, and the probability distribution of all the influencing factors about the load is counted. Therefore, the relevance of each factor and the load prediction is judged again, and the situation that the relevance between some factors and the load prediction is strong, and the relevance between other factors and the load prediction is weak is prevented, particularly the factors which are very close to the relevance coefficient of 0.8 in a test.
In some embodiments, a correlation factor adding module may be further provided for accepting other factors than temperature, humidity, day type, weather, rainfall. Therefore, the diversity of the system on the analysis types of the influencing factors is improved. When new influencing factors appear, the system can react in time. And after the new association factors are added in the association factor adding module, the new association factors and the previous five association factors jointly enter the step two for analyzing the association with the load prediction.
Claims (7)
1. A novel power load prediction influence factor correlation analysis method is characterized by comprising the following steps:
step one, interval division:
acquiring measured data of temperature, humidity, day type, weather, rainfall and load, and performing interval division of the data;
step two, counting the probability distribution of the influence factors about the load:
counting temperature, humidity, day type, weather and rainfallProbability distribution of data with respect to load data, denoted PLi,XjWherein i represents that the load is in the ith load interval; x represents an influence factor, T, W, D, K and R in practical application respectively represent temperature, humidity, day type, weather and rainfall; j indicates that the influence factor is in the jth interval;
step three, calculating a conditional probability set of the influence factors:
on the premise that the load is in the ith interval, the probability that the influence factor X is in the jth interval is calculated, and the calculation formula is
Fourthly, calculating a correlation coefficient:
defining a coefficient of correlation K of the influencing factors to the loadXCalculated as follows:
fifthly, judging the influence factors:
when K isXWhen the load is more than or equal to 0.8, the influence factor X is judged as an effective influence factor of the load prediction and can be used for the load prediction;
when K isXIf the load factor is less than 0.8, the influence factor X is judged as an invalid influence factor of the load prediction and does not need to be considered in the load prediction.
2. The novel power load prediction influence factor correlation analysis method according to claim 1, characterized in that:
in the first step, the partition mode is as follows:
the lowest temperature in the temperature sample data is T1Maximum temperature of T2Temperature is divided into N at equal intervalsTAn i-th interval ofThe right side of the last interval is a closed interval; ith temperatureThe probability of a temperature interval is defined as the ratio of the number of temperature samples belonging to the temperature interval i to all temperature samples, counted as PTi;
The lowest humidity in the humidity sample data is W1Maximum humidity of W2Equally dividing humidity into NWAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith humidity interval is defined as the ratio of the number of humidity samples belonging to the humidity interval i to all the humidity samples, and is counted as PWi;
The date type sample data has two intervals, wherein the interval 1 is a working day and is marked as D1And interval 2 is a non-working day and is marked as D2. The probability of the ith day type interval is defined as the proportion of the number of the day type samples belonging to the day type interval i to all the day type samples, and is counted as PDi;
The weather sample data has 5 sections, wherein the section 1 is sunny, the section 2 is cloudy, the section 3 is cloudy, the section 4 is rain, and the section 5 is snow, which are respectively marked as K1,K2,K3,K4,K5(ii) a The higher the sequence number of the interval is, the higher the priority is, for example, the weather of cloudy to sunny and cloudy to sunny falls into the interval 2; the probability of the ith weather interval is defined as the proportion of the number of the weather samples belonging to the weather interval i to all the weather samples, and is counted as PKi;
The minimum rainfall in the rainfall sample data is R1The maximum rainfall is R2Equally dividing the rainfall into NRAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith rainfall interval is defined as the proportion of the number of rainfall samples belonging to the rainfall interval i to all the rainfall samples, and is counted as PRi;
The lowest load in the load sample data is L1Highest load ofIs L2Will beLoad sampleEqually spaced division into NLAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith load interval is defined as the proportion of the number of load samples belonging to the load interval i to all the load samples, and is counted as PLi。
3. A power load prediction method based on influence factor correlation analysis is characterized by comprising the following steps:
step one, a first data acquisition module acquires the daily electricity consumption data and the daily influence factor change information,
the second data acquisition module acquires power utilization data before the day and influence factor change information before the day;
and step two, analyzing the relevance of the influence factors and the load prediction by an influence factor relevance analysis module, wherein the analysis comprises the following steps:
step 2-1, partition
Acquiring measured data of temperature, humidity, day type, weather, rainfall and load, and completing data interval division according to a method in index parameter definition.
Step 2-2, counting probability distribution of influence factors about load
Counting the probability distribution of temperature, humidity, day type, weather and rainfall data about load data, and recording as PLi,XjWherein i represents that the load is in the ith load interval; x represents an influence factor, T, W, D, K and R in practical application respectively represent temperature, humidity, day type, weather and rainfall; j indicates that the influence factor is in the jth interval. For example, the probability that the temperature is in the 1 st interval of the temperature and the load is in the 1 st interval of the load is PL1,T1。
Step 2-3, calculating a conditional probability set of the influence factors
On the premise that the load is in the ith interval, the probability that the influence factor X is in the jth interval is calculated, and the calculation formula is
Step 2-4, calculating the correlation coefficient
Defining a coefficient of correlation K of the influencing factors to the loadX. Calculated as follows.
Step 2-5, judging influence factors
When K isXWhen the load is more than or equal to 0.8, the influence factor X is judged as an effective influence factor of the load prediction and can be used for the load prediction;
when K isXIf the load is less than 0.8, the influence factor X is judged as an invalid influence factor of the load prediction, and the factors are not required to be considered when the load prediction is carried out in the step three;
and step three, according to the effective influence factors obtained in the step two, the power load prediction module predicts the power load of the remaining time of the day, or predicts the power load of the next day and the future preset time period.
4. The method of claim 3 for power load prediction based on impact factor correlation analysis, wherein:
the partition in step 2-1 is as follows:
the lowest temperature in the temperature sample data is T1Maximum temperature of T2Temperature is divided into N at equal intervalsTAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith temperature interval is defined as the ratio of the number of temperature samples belonging to the temperature interval i to all temperature samples, and is counted as PTi;
The lowest humidity in the humidity sample data is W1Maximum humidity of W2Will wetEqually spaced division into NWAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith humidity interval is defined as the ratio of the number of humidity samples belonging to the humidity interval i to all the humidity samples, and is counted as PWi;
The date type sample data has two intervals, wherein the interval 1 is a working day and is marked as D1And interval 2 is a non-working day and is marked as D2. The probability of the ith day type interval is defined as the proportion of the number of the day type samples belonging to the day type interval i to all the day type samples, and is counted as PDi;
The weather sample data has 5 sections, wherein the section 1 is sunny, the section 2 is cloudy, the section 3 is cloudy, the section 4 is rain, and the section 5 is snow, which are respectively marked as K1,K2,K3,K4,K5(ii) a The higher the sequence number of the interval is, the higher the priority is, for example, the weather of cloudy to sunny and cloudy to sunny falls into the interval 2; the probability of the ith weather interval is defined as the proportion of the number of the weather samples belonging to the weather interval i to all the weather samples, and is counted as PKi;
The minimum rainfall in the rainfall sample data is R1The maximum rainfall is R2Equally dividing the rainfall into NRAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith rainfall interval is defined as the proportion of the number of rainfall samples belonging to the rainfall interval i to all the rainfall samples, and is counted as PRi;
The lowest load in the load sample data is L1The highest load is L2Will beLoad sampleEqually spaced division into NLAn i-th interval ofThe right side of the last interval is a closed interval; the probability of the ith load interval is defined as the proportion of the number of load samples belonging to the load interval i to all the load samples, and is counted as PLi。
5. The method of claim 3 for power load prediction based on impact factor correlation analysis, wherein:
and the system also comprises an associated factor adding module for receiving other factors besides temperature, humidity, day type, weather and rainfall.
6. The method of claim 5 for power load prediction based on impact factor correlation analysis, wherein:
and the newly added correlation factors in the correlation factor adding module enter the step two to analyze the correlation with the load prediction.
7. The method of claim 3 for power load prediction based on impact factor correlation analysis, wherein:
and when the data before the current day collected in the second data collection module is accumulated to a preset time, the second step is carried out again, and the probability distribution of all the influencing factors about the load is counted.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111255136.3A CN113887831A (en) | 2021-10-27 | 2021-10-27 | Novel power load prediction influence factor correlation analysis method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111255136.3A CN113887831A (en) | 2021-10-27 | 2021-10-27 | Novel power load prediction influence factor correlation analysis method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113887831A true CN113887831A (en) | 2022-01-04 |
Family
ID=79013740
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111255136.3A Pending CN113887831A (en) | 2021-10-27 | 2021-10-27 | Novel power load prediction influence factor correlation analysis method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113887831A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117272121A (en) * | 2023-11-21 | 2023-12-22 | 江苏米特物联网科技有限公司 | Hotel load influence factor quantitative analysis method based on Deep SHAP |
-
2021
- 2021-10-27 CN CN202111255136.3A patent/CN113887831A/en active Pending
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117272121A (en) * | 2023-11-21 | 2023-12-22 | 江苏米特物联网科技有限公司 | Hotel load influence factor quantitative analysis method based on Deep SHAP |
CN117272121B (en) * | 2023-11-21 | 2024-03-12 | 江苏米特物联网科技有限公司 | Hotel load influence factor quantitative analysis method based on Deep SHAP |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113962364B (en) | Multi-factor power load prediction method based on deep learning | |
CN116646933A (en) | Big data-based power load scheduling method and system | |
CN111932402A (en) | Short-term power load bidirectional combined prediction method based on similar day and LSTM | |
CN102509173B (en) | A kind of based on markovian power system load Accurate Prediction method | |
CN104598986A (en) | Big data based power load prediction method | |
CN113743673B (en) | Power load prediction method during typhoon | |
CN111915083B (en) | Wind power prediction method and prediction system based on time layered combination | |
CN104598985A (en) | Power load forecasting method | |
CN113554466A (en) | Short-term power consumption prediction model construction method, prediction method and device | |
CN112884008A (en) | Prediction evaluation method and device for running state of power information acquisition system | |
CN112417627A (en) | Power distribution network operation reliability analysis method based on four-dimensional index system | |
CN113887831A (en) | Novel power load prediction influence factor correlation analysis method | |
CN108346009A (en) | A kind of power generation configuration method and device based on user model self study | |
CN111815039A (en) | Weekly scale wind power probability prediction method and system based on weather classification | |
CN114186733A (en) | Short-term load prediction method and device | |
Al-Ja’afreh et al. | An enhanced CNN-LSTM based multi-stage framework for PV and load short-term forecasting: DSO scenarios | |
CN113887809A (en) | Power distribution network supply and demand balance method, system, medium and computing equipment under double-carbon target | |
CN112418662A (en) | Power distribution network operation reliability analysis method using artificial neural network | |
CN116894687A (en) | Power consumption analysis method and system based on machine learning and electronic equipment | |
CN111861587A (en) | System and method for analyzing residential electricity consumption behavior based on hidden Markov model and forward algorithm | |
CN116865327A (en) | Method and system for determining electric balance proportion of new energy incorporation | |
CN115423146A (en) | Self-adaptive runoff forecasting method based on multi-factor nearest neighbor sampling regression and support vector machine | |
CN112001551B (en) | Ground and commercial power grid sales electricity quantity prediction method based on large-user electricity quantity information | |
CN117851908B (en) | Improved on-line low-voltage transformer area electric energy meter misalignment monitoring method and device | |
Huang et al. | Short-term load forecasting method based on heterogeneous big data of power |
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 |