A method and system for predicting gas consumption based on big data analysis
Through big data analysis, divide user types, build sequences, and obtain characteristic values of fluctuating factors, determine the smoothing coefficient, and combine the index smoothing method to solve the problem of inaccurate gas usage prediction and achieve more accurate gas usage prediction.
Patent Information
- Application Number
- CN202411110554.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-14
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-08-14
AI Technical Summary
In the existing gas usage prediction methods, the inappropriate value of the smoothing coefficient leads to inaccurate prediction, making it difficult to effectively respond to short-term fluctuations and changes in gas usage.
Through big data analysis, we divide the original users and new users, build a median gas consumption sequence and a representative user sequence, obtain the characteristic values of representative volatility and volatility factors, determine the appropriate smoothing coefficient, and combine the exponential smoothing method to predict the gas consumption.
It improves the accuracy of gas usage prediction, can respond to fluctuations in applied volume data faster, approaching the actual usage, and solves the problem of inappropriate smoothing coefficient value.
Smart Images

Figure CN118644279B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of gas prediction, and in particular to a gas consumption prediction method and system based on big data analysis. Background Art
[0002] With the acceleration of urbanization and the improvement of residents' living standards, the demand for gas has shown a rapid growth trend, and the annual gas demand has increased to varying degrees. The gas used by residential users and non-residential users is supplied by local gas companies. Gas companies will purchase annual gas quotas before residential users and non-residential users use them to ensure local demand. However, since the annual gas quota is determined in advance before use, there is often a difference between the annual gas quota and the actual gas usage. When the annual gas quota is significantly less than the actual gas usage, it will lead to insufficient supply during peak gas usage; when the annual gas quota is significantly greater than the actual gas usage, it will lead to substandard gas usage, putting the gas company in a situation where it needs to pay a breach of contract fee. Therefore, it is necessary to predict the actual gas usage.
[0003] Generally, the exponential smoothing method can be used to predict the actual gas usage. However, the use of the exponential smoothing method for data prediction requires manual setting of the smoothing coefficient. The value range of the smoothing coefficient is greater than or equal to 0 and less than or equal to 1. Within the value range, when the value of the smoothing coefficient is larger, the predicted value is more sensitive to the data closer in time, and can respond to the change of data faster, but it is also more susceptible to the accidental fluctuations and noise of the adjacent data; within the value range, when the value of the smoothing coefficient is smaller, the obtained predicted value is smoother than the actual collected value, and the obtained predicted value is more in line with the development trend of the data, but the predicted value is also more likely to ignore the short-term volatility of the data. Therefore, it is necessary to determine the value of the smoothing coefficient based on the actual value of the gas usage used for prediction. Summary of the invention
[0004] The present invention provides a gas consumption prediction method and system based on big data analysis to solve the problem of inaccurate gas consumption prediction caused by inappropriate smoothing coefficient value. The technical solution adopted is as follows:
[0005] In a first aspect, an embodiment of the present invention provides a method for predicting gas usage based on big data analysis, the method comprising the following steps:
[0006] Collect gas usage data of all users of the gas company at each sampling time in a local time period;
[0007] Use the big data processing platform to process all gas usage data and obtain gas usage forecast results, including:
[0008] According to the values of the gas usage data of the same user at all consecutive sampling moments, the original users and the newly added users are divided; according to all the gas usage data of the same user, the gas usage sequence of each original user and each newly added user is obtained;
[0009] A median gas usage sequence is constructed based on the data distribution of the gas usage data of all original users at all sampling moments. A representative user sequence of a representative user is determined based on the difference between the median gas usage sequence and the gas usage sequence of the original user. A representative volatility of a representative user is obtained based on the difference between the median gas usage sequence and the representative user sequence of the representative user and the data distribution of the gas usage data contained in the representative user sequence of the representative user. A volatility factor characteristic value of the gas company is obtained by combining the number of newly added users with the number of all users of the gas company.
[0010] The smoothing coefficient value is determined according to the characteristic value of the gas company's fluctuation factor, and the gas consumption forecast result is obtained by combining the gas consumption data of all users.
[0011] Further, the method of dividing the original users and the newly added users according to the values of the gas usage data of the same user at all consecutive sampling moments; and obtaining the gas usage sequence of each original user and each newly added user according to all the gas usage data of the same user includes the following specific methods:
[0012] Arrange all gas usage data of the same user in the order of sampling time to obtain the user's gas usage sequence;
[0013] When the gas usage sequence contains consecutive preset parameter numbers 0, the user corresponding to the gas usage sequence is recorded as a new user, and the user who is not a new user is recorded as an original user.
[0014] Furthermore, the specific method for constructing the median gas consumption sequence is as follows:
[0015] The median of the gas usage data of all original users at the same sampling time is taken as the representative gas usage at the sampling time, and the representative gas usage at all sampling times is arranged in the order of the sampling times to obtain the median gas usage sequence.
[0016] Furthermore, the method of determining the representative user sequence of the representative user according to the difference between the median gas usage sequence and the gas usage sequence of the original user includes the following specific methods:
[0017] Calculate the difference between the gas usage sequence of each original user and the median gas usage sequence, and record the gas usage sequence of the original user corresponding to the smallest difference as the first representative user sequence;
[0018] The gas usage sequences of all original users are clustered to obtain clusters, and all gas usage sequences in the same cluster as the first representative user sequence are recorded as representative user sequences of the original users, and the original users corresponding to the representative user sequences are recorded as representative users.
[0019] Furthermore, the representative volatility of the representative user is obtained by:
[0020] The information entropy of all gas usage data contained in the representative user sequence of the representative user is recorded as the volatility of the representative user, the difference between the median gas usage sequence and the representative user sequence of the representative user is recorded as the volatility difference of the representative user, and the product of the volatility of the representative user and the volatility difference is recorded as the representative volatility of the representative user.
[0021] Furthermore, the characteristic value of the fluctuation factor of the gas company includes the following specific methods:
[0022] Obtain the overall gas volatility of the gas company according to the representative volatility of all representative users;
[0023] The ratio of the number of new users to the number of all users of the gas company is recorded as the user growth rate of the gas company;
[0024] The linear normalized value of the product of the gas company's user growth rate and the overall gas volatility is recorded as the gas company's volatility factor characteristic value.
[0025] Furthermore, the overall gas volatility of the gas company is obtained according to the representative volatility of all representative users, including the specific method of:
[0026] The sum of the representative volatility of all representative users is recorded as the overall gas volatility of the gas company.
[0027] Furthermore, the method of determining the value of the smoothing coefficient according to the characteristic value of the fluctuation factor of the gas company and obtaining the gas consumption prediction result in combination with the gas consumption data of all users includes the following specific methods:
[0028] The sum of the gas usage data of all users at the same sampling time is taken as the total gas usage at the sampling time, and the total gas usage at all sampling times is arranged in the order of the sampling times to obtain a total gas usage sequence;
[0029] The gas consumption forecast result is obtained based on the total gas consumption sequence and the characteristic values of the gas company's fluctuation factors.
[0030] Furthermore, the gas consumption forecast result is obtained according to the total gas consumption sequence and the characteristic value of the gas company's fluctuation factor, including the specific method of:
[0031] The characteristic value of the gas company's volatility factor is used as the smoothing coefficient of the exponential smoothing method. The exponential smoothing method is used to predict the total gas usage series to obtain the gas usage prediction results.
[0032] In a second aspect, an embodiment of the present invention further provides a gas usage prediction system based on big data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above methods when executing the computer program.
[0033] The beneficial effects of the present invention are:
[0034] The present application first divides original users and new users according to the values of the gas usage data of the users at multiple sampling moments in a local time period. New users are users who have not used gas for a long time. The gas usage habits of these users who have not used gas for a long time will affect the analysis of the gas usage habits of all users of the gas company. Therefore, the new users are divided out and the gas usage habits of these users are analyzed separately. The original users are users who have not used gas for a long time. According to the different gas usage habits of different users and the different gas usage habits of the same user at different times, the present application first screens out the representative user sequence of the user according to the gas usage data of the original users. The representative user sequence of the user is the gas usage data that can represent the gas usage characteristics of the user. According to the difference between the median gas usage sequence and the representative user sequence of the representative user, and the data distribution of the gas usage data contained in the representative user sequence of the representative user, the gas usage data of the representative user is evaluated. The fluctuation degree of gas consumption data is calculated to obtain the representative volatility of the representative user; then, combined with the characteristic that the gas consumption data of the newly added users will increase the annual gas consumption, the impact of the gas consumption data of the newly added users on the overall fluctuation of the gas consumption data of all users of the gas company is considered, and the number of newly added users and the number of all users of the gas company are combined to obtain the characteristic value of the fluctuation factor of the gas company. When the characteristic value of the fluctuation factor of the gas company is larger, the volatility of the gas consumption data of all users of the gas company is greater, and a larger smoothing coefficient should be selected to make the predicted value more sensitive to the data closer in time, so that the obtained predicted value can respond to the fluctuation of the gas consumption data faster, and the obtained predicted value is closer to the actual gas consumption; finally, the value of the smoothing coefficient is determined according to the characteristic value of the fluctuation factor of the gas company, and the gas consumption data of all users are combined to obtain the gas consumption prediction result based on big data analysis, so as to solve the problem of inaccurate gas consumption prediction caused by inappropriate smoothing coefficient value. BRIEF DESCRIPTION OF THE DRAWINGS
[0035] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.
[0036] Figure 1 A schematic flow chart of a method for predicting gas consumption based on big data analysis provided by an embodiment of the present invention;
[0037] Figure 2 A flow chart of obtaining characteristic values of fluctuation factors provided by one embodiment of the present invention. DETAILED DESCRIPTION
[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0039] See also Figure 1 , which shows a flow chart of a method for predicting gas usage based on big data analysis provided by an embodiment of the present invention, the method comprising the following steps:
[0040] Step S001, collect the gas usage data of all users of the gas company at each sampling time in a local time period, use the big data processing platform to process all the gas usage data, and divide the original users and the newly added users according to the values of the gas usage data of all consecutive sampling times of the same user; according to all the gas usage data of the same user, obtain the gas usage sequence of each original user and each newly added user.
[0041] Gas usage data is collected from all users of gas companies that need to make annual gas usage quota forecasts.
[0042] Preferably, in this embodiment, the collection period of gas usage data for each user is five years, and the gas usage data is collected once every natural month. In actual application, as other implementation methods, the implementer can decide the collection period of gas usage data for each user, the time interval for collecting gas usage data, and the sampling time of gas usage data according to actual conditions, and this application does not impose any special restrictions.
[0043] Since the amount of gas usage data collected is huge, the big data processing platform Apache Hadoop is used to process the gas usage data to improve the processing and storage speed of the gas usage data.
[0044] The user's gas usage data is cleaned, outliers in the gas usage data are removed, missing values in the gas usage data are filled, and the gas usage data is dimensioned. In this embodiment, the LOF outlier detection algorithm is used to identify outliers in the gas usage data and the mean filling method is used to fill the outliers and missing values. The Z-Score standard normalization method is used to de-dimensionalize the gas usage data. The LOF outlier detection algorithm identifies outliers, the mean filling method is used to fill data, and the Z-Score standard normalization method is used to de-dimensionalize the gas usage data. These are all well-known technologies and will not be described in detail. In actual application, as other implementation methods, on the basis of achieving the purpose of filling out outliers and missing values, the implementer may also use other existing methods such as median filling, mode filling, random sampling filling, K-nearest neighbor filling, regression filling, etc. to fill out outliers and missing values. This application does not impose any special restrictions.
[0045] All gas usage data of the same user are arranged in the order of sampling time to obtain the user's gas usage sequence.
[0046] When the gas usage sequence contains consecutive preset parameter numbers 0, the user corresponding to the gas usage sequence is recorded as a new user, and the user who is not a new user is recorded as an original user.
[0047] In this embodiment, the value of the preset parameter is 6. It is understandable that the newly added users are users who do not use gas for a long time. The gas usage habits of these users who do not use gas for a long time will affect the analysis of the gas usage habits of all users of the gas company. Therefore, the newly added users are divided out and the gas usage habits of these users are analyzed separately.
[0048] At this point, the gas usage sequences of the original users and the newly added users are obtained.
[0049] Step S002, constructing a median gas usage sequence according to the data distribution of the gas usage data of all original users at all sampling moments, determining a representative user sequence of the representative user according to the difference between the median gas usage sequence and the gas usage sequence of the original users, obtaining the representative volatility of the representative user according to the difference between the median gas usage sequence and the representative user sequence of the representative user, and the data distribution of the gas usage data contained in the representative user sequence of the representative user, and obtaining the volatility factor characteristic value of the gas company in combination with the number of newly added users and the number of all users of the gas company.
[0050] Different users have different habits of using gas, and the same user has different habits of using gas at different times. For example, some users are busy at work and prefer convenient and quick meals, so they use less gas for cooking. During long holidays, they occasionally consider meals that require a long cooking time, and the amount of gas used occasionally increases; but some users are used to enjoying meals that require a long cooking time, so they use more gas for cooking, but they also occasionally consider relaxing for a while, choosing to eat out and not using gas at home; some restaurants' main business is wedding banquets and celebrations, and these restaurants use more gas in months with more long holidays such as May and October. Therefore, the fluctuation trend of gas usage data in the gas usage sequence of the same user will be different, and the fluctuation of gas usage data in the gas usage sequence of different users will be different.
[0051] Therefore, it is necessary to analyze the overall fluctuation of the gas usage data of all users of the gas company according to the fluctuation of the gas usage data in the gas usage sequence of the original users, and determine the appropriate smoothing coefficient. When the overall fluctuation is large, a larger smoothing coefficient should be selected, and when the overall fluctuation is small, a smaller smoothing coefficient should be selected to make the forecast result of the annual gas consumption quota closer to the sum of the actual consumption of the gas usage data of all users.
[0052] Firstly, the gas usage data that can represent the gas usage characteristics of the users are screened out based on the gas usage data of all original users.
[0053] Specifically, the median of the gas usage data of all original users in the same natural month is taken as the representative gas usage of the natural month, and the representative gas usage of all natural months is arranged in the order of sampling time to obtain the median gas usage sequence.
[0054] It is understandable that the median is used to measure the central tendency of data. Although it may not be possible to find an original user's gas usage sequence that is exactly the same as the median gas usage sequence, that is, no original user's gas usage data can constitute the median gas usage sequence, but the gas usage data that constitutes the median gas usage sequence is the data value of all gas usage data sets at each sampling time.
[0055] The representative user sequence is determined according to the difference between the median gas usage sequence and the gas usage sequence of the original user.
[0056] The difference between the gas usage sequence of each original user and the median gas usage sequence is calculated, and the gas usage sequence of the original user corresponding to the smallest difference is recorded as the first representative user sequence.
[0057] The K-Means clustering algorithm is used to cluster the gas usage sequences of all original users to obtain cluster clusters. All gas usage sequences in the same cluster cluster as the first representative user sequence are recorded as representative user sequences of the original users, and the original users corresponding to the representative user sequences are recorded as representative users.
[0058] This embodiment uses the K-Means clustering algorithm to cluster the gas usage sequences of all original users. In actual application, as other implementation methods, on the basis of achieving the purpose of clustering the gas usage sequences of all original users, the implementer can use other existing methods such as the OPTICS clustering algorithm, the DBSCAN clustering algorithm, etc. for clustering, and this application does not make any special restrictions.
[0059] The representative volatility of the representative user is obtained according to the difference between the median gas usage sequence and the representative user sequence of the representative user, and the data distribution of the gas usage data contained in the representative user sequence of the representative user.
[0060] Preferably, as an embodiment of the present application, the information entropy of all gas usage data contained in the representative user sequence of the representative user is recorded as the volatility of the representative user, the difference between the median gas usage sequence and the representative user sequence of the representative user is recorded as the volatility difference of the representative user, and the product of the volatility of the representative user and the volatility difference is recorded as the representative volatility of the representative user.
[0061] Among them, this embodiment selects DTW distance as the difference measure between the gas usage sequence of the original user and the median gas usage sequence, and selects DTW distance as the difference measure between the median gas usage sequence and the representative user sequence of the representative user. Calculating the DTW distance of the two sequences is a well-known technology and will not be repeated. As other implementation methods, the implementer can use other existing methods such as JS divergence, MASS (Mueen's Algorithm for Similarity Search) similarity search distance, etc. to measure the difference between the gas usage sequence of the original user and the median gas usage sequence, and the difference between the median gas usage sequence and the representative user sequence of the representative user, and this application does not make special restrictions.
[0062] When the data distribution of the gas usage data contained in the representative user sequence of the representative user is more chaotic, and the difference between the median gas usage sequence and the representative user sequence of the representative user is greater, the representative volatility of the representative user is greater. At this time, the overall volatility of the gas usage data of all users of the gas company is larger.
[0063] The overall gas volatility of the gas company is obtained according to the representative volatility of all representative users. The overall gas volatility of the gas company is positively correlated with the representative volatility of all representative users.
[0064] Preferably, as an embodiment of the present application, the sum of the representative volatility of all representative users is recorded as the overall gas volatility of the gas company.
[0065] When the representative volatility of all representative users is greater, the overall volatility of gas is greater. At this time, the overall volatility of gas usage data of all users of the gas company is larger.
[0066] When the gas usage data of newly added users is directly used to analyze the gas usage of all users of the gas company, since the newly added users are users who have not used gas for a long time, the gas usage data of these newly added users will affect the prediction of the annual gas usage quota. However, if the annual gas usage quota is predicted only based on the gas usage of the original users, these newly added users will also use gas, so the predicted value obtained must be smaller than the actual value. Therefore, further consider the impact of the gas usage data of newly added users on the overall fluctuation of the gas usage data of all users of the gas company.
[0067] The more new users there are compared to the total number of users of the gas company, the more gas used by these new users will increase the gas usage of all users, and the more the annual gas consumption quota that needs to be predicted for the gas company should increase.
[0068] Obtain the user growth rate of the gas company based on the number of new users and the number of all users of the gas company.
[0069] Preferably, as an embodiment of the present application, the ratio of the number of new users to the number of all users of the gas company is recorded as the user growth rate of the gas company.
[0070] Among them, the calculation formula for the user growth rate of the gas company is:
[0071]
[0072] In the formula, represents the user growth rate of the gas company; Indicates the number of newly added users; Represents the number of all customers of the gas company.
[0073] According to the user growth rate and overall gas volatility of the gas company, the fluctuation factor characteristic value of the gas company is obtained. The fluctuation factor characteristic value of the gas company is positively correlated with the user growth rate and overall gas volatility of the gas company.
[0074] Preferably, as an embodiment of the present application, a linear normalized value of the product of the user growth rate of the gas company and the overall volatility of the gas is recorded as the volatility factor characteristic value of the gas company.
[0075] Among them, the calculation formula of the characteristic value of the fluctuation factor of the gas company is:
[0076]
[0077] In the formula, represents the characteristic value of the fluctuation factor of the gas company; represents the user growth rate of the gas company; Indicates the overall volatility of gas from gas companies; represents the linear normalization function.
[0078] When the user growth rate and overall gas volatility of a gas company are greater, the characteristic value of the gas company's volatility factor is greater. At this time, the volatility of the gas usage data of all users of the gas company is greater, and a larger smoothing coefficient should be selected to make the predicted value more sensitive to data closer in time, so that the obtained predicted value can respond to the fluctuation of the gas usage data more quickly and make the obtained predicted value closer to the actual gas usage.
[0079] At this point, the characteristic value of the fluctuation factor of the gas company is obtained. The flow chart of obtaining the characteristic value of the fluctuation factor is as follows: Figure 2 shown.
[0080] Step S003, determining the value of the smoothing coefficient according to the characteristic value of the fluctuation factor of the gas company, and obtaining the gas usage prediction result in combination with the gas usage data of all users.
[0081] The sum of the gas usage data of all users at the same sampling time is taken as the total gas usage at the sampling time, and the total gas usage at all sampling times is arranged in the order of the sampling times to obtain a total gas usage sequence.
[0082] The characteristic value of the gas company's volatility factor is used as the smoothing coefficient of the exponential smoothing method. The exponential smoothing method is used to predict the total gas usage series to obtain the annual gas consumption quota forecast value of the gas company. The annual gas consumption quota forecast value of the gas company is the gas consumption forecast result.
[0083] At this point, the big data processing platform Apache Hadoop is used to obtain the annual gas consumption quota forecast value of the gas company and obtain the gas consumption forecast result based on big data analysis.
[0084] Based on the same inventive concept as the above method, an embodiment of the present invention also provides a gas usage prediction system based on big data analysis, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above-mentioned gas usage prediction methods based on big data analysis.
[0085] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A gas consumption prediction method based on big data analysis, characterized in that: The method comprises the following steps: Collect gas usage data of all users of the gas company at each sampling time in a local time period; Use the big data processing platform to process all gas usage data and obtain gas usage forecast results, including: According to the values of the gas usage data of the same user at all consecutive sampling moments, the original users and the newly added users are divided; according to all the gas usage data of the same user, the gas usage sequence of each original user and each newly added user is obtained; A median gas usage sequence is constructed based on the data distribution of the gas usage data of all original users at all sampling moments. A representative user sequence of a representative user is determined based on the difference between the median gas usage sequence and the gas usage sequence of the original user. A representative volatility of a representative user is obtained based on the difference between the median gas usage sequence and the representative user sequence of the representative user and the data distribution of the gas usage data contained in the representative user sequence of the representative user. A volatility factor characteristic value of the gas company is obtained by combining the number of newly added users with the number of all users of the gas company. The smoothing coefficient value is determined according to the characteristic value of the gas company's fluctuation factor, and the gas consumption forecast result is obtained by combining the gas consumption data of all users; The method of determining the representative user sequence of the representative user according to the difference between the median gas usage sequence and the gas usage sequence of the original user includes the following specific methods: Calculate the difference between the gas usage sequence of each original user and the median gas usage sequence, and record the gas usage sequence of the original user corresponding to the smallest difference as the first representative user sequence; Clustering the gas usage sequences of all original users to obtain cluster clusters, recording all gas usage sequences in the same cluster cluster as the first representative user sequence as representative user sequences of the original users, and recording all original users corresponding to the representative user sequences as representative users; The specific method for obtaining the representative volatility of the representative user is as follows: The information entropy of all gas usage data contained in the representative user sequence of the representative user is recorded as the volatility of the representative user, the difference between the median gas usage sequence and the representative user sequence of the representative user is recorded as the volatility difference of the representative user, and the product of the volatility of the representative user and the volatility difference is recorded as the representative volatility of the representative user; The specific method of the characteristic value of the fluctuation factor of the gas company includes: The sum of the representative volatility of all representative users is recorded as the overall gas volatility of the gas company; The ratio of the number of new users to the number of all users of the gas company is recorded as the user growth rate of the gas company; The linear normalized value of the product of the gas company's user growth rate and the overall gas volatility is recorded as the gas company's volatility factor characteristic value.
2. A method for predicting gas consumption based on big data analysis according to claim 1, characterized in that: The method of dividing the original users and the newly added users according to the values of the gas usage data of the same user at all consecutive sampling moments; and obtaining the gas usage sequence of each original user and each newly added user according to all the gas usage data of the same user includes the following specific methods: Arrange all gas usage data of the same user in the order of sampling time to obtain the user's gas usage sequence; When the gas usage sequence contains consecutive preset parameter numbers 0, the user corresponding to the gas usage sequence is recorded as a new user, and the user who is not a new user is recorded as an original user.
3. A method for predicting gas consumption based on big data analysis according to claim 1, characterized in that: The specific method of constructing the median gas consumption sequence is as follows: The median of the gas usage data of all original users at the same sampling time is taken as the representative gas usage at the sampling time, and the representative gas usage at all sampling times is arranged in the order of the sampling times to obtain the median gas usage sequence.
4. A method for predicting gas consumption based on big data analysis according to claim 1, characterized in that: The method of determining the smoothing coefficient value according to the characteristic value of the gas company's fluctuation factor and obtaining the gas consumption prediction result in combination with the gas consumption data of all users includes: The sum of the gas usage data of all users at the same sampling time is taken as the total gas usage at the sampling time, and the total gas usage at all sampling times is arranged in the order of the sampling times to obtain a total gas usage sequence; The gas consumption forecast result is obtained based on the total gas consumption sequence and the characteristic values of the gas company's fluctuation factors.
5. A method for predicting gas consumption based on big data analysis according to claim 4, characterized in that: The specific method for obtaining the gas consumption prediction result according to the total gas consumption sequence and the characteristic value of the gas company's fluctuation factor is as follows: The characteristic value of the gas company's volatility factor is used as the smoothing coefficient of the exponential smoothing method. The exponential smoothing method is used to predict the total gas usage series to obtain the gas usage prediction results.
6. A gas consumption prediction system based on big data analysis, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 5 are implemented.
