Winter city gas daily load prediction method based on human body comfort index similarity

Abstract

The invention relates to a winter urban gas daily load prediction method based on human body comfort index similarity. The method comprises the following steps: a, taking winter historical data including human body comfort indexes and gas daily loads in recent years of a prediction region; b, according to a trend similarity formula of a human comfort index curve, finding out two time periods H and I which have the maximum similarity with the change trend of the human comfort index curve in the prediction time period J and have the second similarity with the change trend of the human comfort index curve in the prediction time period J; c, according to the historical data of the time period H and the time period I, the day-by-day load-comfort change rates of the days corresponding to the time period H and the time period I are solved; and d, calculating the corresponding gas load of the jth day in the prediction time period J. And e, performing analogy according to the steps c and d, and solving the gas load of k days after the jth day in the prediction time period J. The method has the advantages of being simple and convenient in calculation process, high in precision, stable in result and the like, and provides a basis for gas volume scheduling and peak regulation of urban gas companies.

Description

Technical Field

[0001] The invention relates to a method for predicting daily urban gas load in winter based on similarity of human comfort index, and belongs to the technical field of gas load prediction. Background Art

[0002] Gas load refers to gas consumption, which can be categorized as industrial, residential, and commercial loads based on gas usage structure. Load forecasts can be categorized into short-term, medium-term, and long-term forecasts based on time periods. Short-term city gas load forecasting is crucial for effectively guiding gas storage, peak load regulation, and scheduling within the gas supply system. It is influenced by factors such as gas usage structure, date type, and meteorological factors. Since industrial and commercial gas consumption is relatively stable, it is less affected by weather, whereas residential load is significantly impacted by meteorological factors such as temperature, wind speed, and humidity. Therefore, in cities where industrial and commercial users make up a small proportion and residential users predominate, meteorological factors are considered key influencing factors influencing daily city gas consumption, making them crucial for short-term winter gas load forecasting.

[0003] The impact of meteorological factors on short-term loads is attracting widespread attention from scholars both in China and abroad. Currently, winter short-term gas load forecasting methods primarily focus on the relationship between raw meteorological parameters such as temperature and solar radiation and historical gas load data, using curve fitting and artificial neural networks to develop prediction models. However, the impact of daily winter city gas load is often a result of the combined effects of meteorological factors such as temperature, humidity, and wind speed. Trend forecasts that only consider meteorological factors like temperature can result in significant errors. Summary of the Invention

[0004] The purpose of this invention is to address the defects of the existing gas daily load prediction method and propose a winter city gas daily load prediction method based on the similarity of human comfort index. It can combine the weather forecast to make a more accurate prediction of the gas daily load in the next week, providing a scientific basis for gas company's gas volume scheduling.

[0005] The technical solution provided by the present invention is: a method for predicting daily city gas load in winter based on similarity of human comfort index, comprising the following steps: Step a: Obtain winter historical data of the forecast area in recent years, including human comfort index and gas daily load. X ; Step b: According to the trend similarity formula of the human comfort index curve, find the time period with the prediction time J The two time periods with the largest and second largest similarity in the change trend of the human comfort index curve H and I ; Step c: According to the time period H and time period IThe historical data of K hi =（L h -L i ） / （S h -S i ） Find the time period H and time period I Daily load-comfort change rate of the corresponding day K hi ; In the formula, the two time periods with similar trends in human comfort index are H and I The first day of h day and day i day, L h For the h Daily gas load, L i For the i Daily gas load, S h For the h Human comfort index of the day, S i For the i Human comfort index of the day; Step d: According to the formula L j =K hi · （S j -S i ）+L i Calculate the forecast time period J The corresponding j Daily gas load L j ; Where, the forecast time period J The first day is recorded as J day, L j Forecast period J Middle j Daily gas load.

[0006] Step e: Follow steps c and d and calculate the predicted time period. J Middle j Day after k Daily gas load.

[0007] The method for predicting daily urban gas load in winter based on similar human comfort index, the prediction time period J Middle j Daily human comfort index S j , time period I Middle i Day and time period H Middle h Daily human comfort index S i 、 S h Obtained by the following calculation method: (1) Where: S ssd for S h or S i or S j , T is the average temperature of the corresponding date, ℃; R h is the average relative humidity on the corresponding day, %; is the average wind speed on the corresponding date, m / s; forecast time period J Middle j The average temperature, average relative humidity, and average wind speed of the day can be obtained from the local weather forecast; the average wind speed for the corresponding date is m / s. Forecast period J Middle j The daily average temperature, average relative humidity, and average wind speed can be obtained from the local weather forecast.

[0008] The method for predicting the daily load of winter city gas based on similar human comfort index, the time period H , time period I and time period J It is 2-8 days and the number of days is the same.

[0009] In the winter city gas daily load forecasting method based on the similarity of the human comfort index, the similarity of the change trend of the human comfort index curve is obtained by the following method: The two time periods I 、 J The first day is recorded as i day and day j day, no. i day and day j Day and after k The daily human comfort index change series are S i =（Si ，S i+1 ，…，S i+k ） and S j =（S j ，S j+1 ，…，S j+k ） , S i and S j Calculated according to formula (1); then the similarity of the human comfort index of these two time periods is H ij for: (2) Where, H ij The value range of is 0~1, and the closer it is to 1, the greater the similarity of the curve trend; according to the short-term gas load change law and the accuracy of weather forecast, k The value is not greater than 7.

[0010] The method for predicting the daily load of urban gas in winter based on the similarity of human comfort index includes the winter historical data of human comfort index and daily gas load. X The normal working day and holiday data are divided into two groups, respectively R 1 、 R 2 ;like J For holiday periods, step b is to R 2 Find the time period that matches the forecast period J The two time periods with the largest and second largest similarity in the change trend of the human comfort index curve I and H ;like J For normal working day period, step b is to group R 1 Find the time period that matches the forecast period J The two time periods with the largest and second largest similarity in the change trend of the human comfort index curve I and H .

[0011] The present invention provides a short-term winter gas load forecast for cities dominated by residential users. This forecast uses daily gas load and historical meteorological data as a basis, comprehensively considering the impact of meteorological factors such as temperature, humidity, and wind speed on the daily winter city gas load. This method boasts a simple calculation process, high accuracy, and stable results. It provides a basis for city gas companies to schedule and regulate gas volumes, and is also crucial for guiding the classification of different cities based on their gas usage structure and developing corresponding forecasting models. DETAILED DESCRIPTION

[0012] Specific embodiment: Taking a certain region in China as an example, the content and implementation of the present invention are explained: The technical solution of this invention is to study the similarity of human comfort index curve shapes, provide a definition of similar human comfort index days, and provide a method for calculating their similarity. Based on the daily gas load difference between similar human comfort index days, the daily load-comfort index change rate is calculated. Combining this change rate with the similar human comfort index days, the corresponding daily gas load value is calculated.

[0013] The specific process is as follows: 1. Taking a certain region in China as an example, we take the historical winter data (human comfort index and daily gas load) for the forecast region over the past two years and separate the data for normal working days and holidays into two groups. Assuming that January 21 to 25, 2023, are the forecast days and all are normal working days, Table 1 shows some weather information for this period. The human comfort index for this period is obtained using formula (1). Table 1 Partial weather information for a certain area on a forecast day

[0014] date Average temperature / (℃) Relative humidity / % Wind speed / (m / s) Human comfort index 2023.1.21 8 78 2.5 13.82 2023.1.22 7 82 2.5 12.21 2023.1.23 7.5 79 3 13.02 2023.1.24 8.5 80 3 14.14 2023.1.25 8.5 80 2.5 14.62 According to formula (2), the human comfort index similarity calculation formula is calculated. From the historical data of normal working days, the two time periods with the greatest similarity to the comfort index curve change trend of the forecast day time period are found, namely December 10 to 14, 2022 and January 14 to 18, 2023. Table 2 shows the historical data of this time period. Table 2 Historical data of comfort index trends in similar periods

[0015] date Human comfort index Daily gas load / (m³) 2022.12.10 16.22 244733 2022.12.11 13.82 242073 2022.12.12 15.32 245726 2022.12.13 15.42 243322 2022.12.14 15.41 243444 2023.1.14 10.61 234847 2023.1.15 9.01 236465 2023.1.16 12.21 230000 2023.1.17 11.41 230114 2023.1.18 12.18 231345 3. Based on the known daily gas load and human comfort index data in Table 2, calculate the daily load-comfort index change rate for these two time periods according to formula (3): K hi The load-comfort index change rate between the forecast day and the day with the maximum trend similarity is replaced by this value. The daily load values ​​for the forecast period from January 21 to 25, 2023 are calculated according to formula (4). The results are shown in Table 3. K hi =（L h -L i ） / （S h -S i ） (3) In the formula, the two time periods with similar trends in human comfort index are H and I The first day of h day and day i day, L h For the h Daily gas load, L i For the i Daily gas load, S h For the h Human comfort index of the day, S i For the i Human comfort index of the day; (4) Where j is the forecast day, and h is the similar day with the greatest similarity to the human comfort index of forecast day j. Table 3. Prediction result error statistics

[0016] date Actual load value Forecast load value Error rate / % 2023.1.21 239784 229098 4.456 2023.1.22 241095 232625 3.513 2023.1.23 228551 225932 1.145 2023.1.24 227080 219257 3.445 2023.1.25 227581 221957 2.471 Table 3 compares the prediction results calculated by this method with the actual results. The average error is 3.01%; the maximum error is 4.46%, which occurs on January 21, 2023; the minimum error is 1.15%, which occurs on January 23, indicating that the prediction method is reasonable and reliable.

Claims

1. A winter city gas daily load forecasting method based on similarity of human comfort index, The method is characterized by comprising the following steps: Step a: Obtain winter historical data of the forecast area in recent years, including human comfort index and gas daily load. X ; Step b: Find the time period that matches the forecast J The two time periods with the largest and second largest similarity in the change trend of the human comfort index curve H and I ; Step c: According to the time period H and time period I The historical data of K hi =（L h -L i ） / （S h -S i ） Find the time period H and time period I Daily load-comfort change rate of the corresponding day K hi ; In the formula, the two time periods with similar trends in human comfort index are H and I The first day of h day and day i day, L h For the h Daily gas load, L i For the i Daily gas load, S h For the h Human comfort index of the day, S i For the i Human comfort index of the day; Step d: According to the formula L j =K hi · （S j -S i ）+L i Calculate the forecast time period J The corresponding j Daily gas load L j ; Where, the forecast time period J The first day is recorded as j day, L j Forecast period J Middle j Daily gas load; Step e: Follow steps c and d and calculate the predicted time period. J Middle j Day after k Daily gas load.

2. A winter city gas daily load forecasting method based on human comfort index similarity according to claim 1, characterized in that: The forecast period J Middle j Daily human comfort index S j , time period I Middle i Day and time period H Middle h Daily human comfort index S i 、 S h Obtained by the following calculation method: （1） Where: S ssd for S h or S i or S j , T is the average temperature of the corresponding date, ℃; R h is the average relative humidity on the corresponding day, %; is the average wind speed on the corresponding date, m / s; forecast time period J Middle j The daily average temperature, average relative humidity, and average wind speed can be obtained from the local weather forecast.

3. The method for predicting daily city gas load in winter based on similarity of human comfort index according to claim 2, characterized in that: The time period H , time period I and time period J It is 2-8 days and the number of days is the same.

4. The method for predicting daily city gas load in winter based on similarity of human comfort index according to claim 3, characterized in that: The similarity of the change trend of the human comfort index curve is obtained by the following method: The two time periods I 、 J The first day is recorded as i day and day j day, no. i day and day j Day and after k The daily human comfort index change series are S i =（S i ，S i+1 ，…，S i+k ） and S j =（S j ，S j+1 ，…，S j+k ） , S i and S j Calculated according to formula (1); then the similarity of the human comfort index of these two time periods is H ij for: （2） Where, , H ij The value range of is 0~1, and the closer it is to 1, the greater the similarity of the curve trends; According to the short-term gas load variation law and the accuracy of weather forecast, k The value is not greater than 7.

5. A winter city gas daily load forecasting method based on human comfort index similarity according to any one of claims 1 to 4, characterized in that: Will include winter historical data on human comfort index and daily gas load X The normal working day and holiday data are divided into two groups, denoted as R 1 、R 2 ;like J For holiday period, step b is to 2 Find the time period that matches the forecast period J The two time periods with the largest and second largest similarity in the change trend of the human comfort index curve I and H ; like J For normal working day period, step b is to group R 1 Find the time period that matches the forecast period J The two time periods with the largest and second largest similarity in the change trend of the human comfort index curve I and H .

Images