Storage medium, gas station product oil sales volume prediction method, device and equipment

Abstract

The invention discloses a memory, and a gas station product oil sales volume prediction method, device and equipment. The method comprises the steps of obtaining liquid level data of a gas station liquid level instrument and generating a first array based on a time sequence; sampling elements of the first array according to a preset time period to generate a second array; generating a third array according to the second array, wherein subtraction is carried out on the liquid level data in front and back adjacent elements in the second array to obtain multiple oil consumption data; establishing a corresponding relationship between the oil consumption data and the medium time data of the later element; generating a third array according to the corresponding relationship between the oil consumption data and the time data; determining the oil consumption data with a negative value in the third array as abnormal data, and performing data correction on the abnormal data; and generating an oil consumption prediction model through data training according to the third array. According to the method, the timeliness of modeling data is higher, and the data size is larger, so that an oil consumption prediction model with better prediction precision can be constructed.

Description

Technical Field

[0001] The present invention relates to the field of refined oil sales, and in particular to a method, device and equipment for predicting the sales volume of refined oil at a storage device or a gas station. Background Art

[0002] Finished oil distribution is the last and most widely distributed business node in the finished oil supply chain. Gas station demand forecast has a huge impact on the finished oil supply chain.

[0003] Traditional gas station demand formulation relies on the years of experience accumulated by practitioners, but the relevant experience is difficult to express in an intuitive and quantitative form.

[0004] In the prior art, there are already technical solutions that use machine learning methods to train historical sales data of gas stations in order to formulate and implement personalized needs for gas stations in different regions and with different characteristics under rapidly changing market scenarios.

[0005] The inventor has found through research that the existing method for predicting the sales volume of refined oil products at gas stations using machine learning methods has at least the following defects:

[0006] The forecast error is large and cannot provide effective data support for the distribution of refined oil products and the establishment of refined oil supply chains.

[0007] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0008] The purpose of the present invention is to improve the prediction accuracy of the sales volume of refined oil products at a gas station.

[0009] The present invention provides a method for predicting the sales volume of refined oil at a gas station, comprising the steps of:

[0010] S11, obtaining liquid level data of a liquid level meter at a gas station, and generating a first array based on a time series; the elements of the first array include liquid level data and corresponding time data;

[0011] S12, sampling the elements of the first array according to a preset time period to generate a second array;

[0012] S13, generating a third array according to the second array, comprising: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array respectively to obtain a plurality of oil consumption data; establishing a correspondence between the oil consumption data and the time data in the rear element; generating the third array according to the correspondence between the oil consumption data and the time data;

[0013] S14, determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data;

[0014] S15. Using the elements of the third array as modeling data, an oil consumption prediction model is generated through data training, wherein the oil consumption prediction model is used to predict oil consumption prediction data for a preset number of days in the future.

[0015] Preferably, the present invention further comprises:

[0016] S16. Determine a fuel replenishment plan based on the fuel consumption forecast data for a preset number of days in the future and the current inventory of refined oil products at the gas station.

[0017] In another aspect of the present invention, a device for predicting the sales volume of refined oil at a gas station is provided, comprising:

[0018] A data acquisition unit, used to acquire liquid level data of a liquid level meter at a gas station, and generate a first array based on a time series; the elements of the first array include liquid level data and corresponding time data;

[0019] A data sampling unit, used for sampling the elements of the first array according to a preset time period to generate a second array;

[0020] The modeling data generating unit is used to generate a third array according to the second array, including: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array to obtain a plurality of oil consumption data; establishing a corresponding relationship between the oil consumption data and the time data in the rear element; generating the third array according to the corresponding relationship between the oil consumption data and the time data;

[0021] A modeling data correction unit, used for determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data;

[0022] The model building unit is used to use the elements of the third array as modeling data to generate an oil consumption prediction model through data training, wherein the oil consumption prediction model is used to predict the oil consumption prediction data for a preset number of days in the future.

[0023] Preferably, the present invention further comprises:

[0024] The plan generation unit is used to determine the oil replenishment plan based on the oil consumption forecast data for a preset number of days in the future and the current inventory of refined oil at the gas station.

[0025] On the other hand, there is provided a device for predicting the sales volume of refined oil at a gas station, which includes a computer program stored on a medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the methods described in the above aspects and achieves the same technical effects.

[0026] On the other hand, a storage medium is provided on which a computer program is stored. When the computer program is executed by a processor, each step of the method for predicting the sales volume of refined oil at a gas station as described in any one of the above items is implemented.

[0027] Compared with the prior art, the present invention has the following beneficial effects:

[0028] The present invention no longer uses the sales data of gas stations as modeling data for constructing prediction models, but collects and stores the liquid level data of gas station liquid level meters in real time to generate modeling data with better data granularity and timeliness. When generating modeling data, the present invention obtains the oil consumption data at each time point by calculating the liquid level difference between adjacent data. In this way, after constructing the modeling data consisting of the oil consumption data and the corresponding time, an oil consumption prediction model for predicting the oil consumption prediction data for a preset number of days in the future can be generated through data training.

[0029] It can be seen from the above that, compared with the sales data used in the prior art, the change time points of the oil data in the modeling data of the present invention correspond better to the actual sales time points of the oil; in addition, in the present invention, the collection period of the liquid level data of the gas station level meter is very short, so the amount of data generated is very large, so that sufficient modeling data can be generated according to different sampling periods; that is, the modeling data in the present invention is more timely and the data volume is larger, so it is possible to construct an oil consumption prediction model with better prediction accuracy; this can also provide effective data support for the distribution of refined oil and the establishment of a refined oil supply chain.

[0030] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention and implement it according to the contents of the specification, and to make the above and other purposes, technical features and advantages of the present invention easier to understand, one or more preferred embodiments are listed below and described in detail with the accompanying drawings as follows. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 It is a step diagram of the method for predicting the sales volume of refined oil products at a gas station described in the present invention;

[0032] Figure 2 It is a structural schematic diagram of the device for predicting the sales volume of refined oil products at a gas station of the present invention;

[0033] Figure 3 It is a structural schematic diagram of the device for predicting the sales volume of refined oil products at a gas station described in the present invention. DETAILED DESCRIPTION

[0034] The specific implementation modes of the present invention are described in detail below in conjunction with the accompanying drawings, but it should be understood that the protection scope of the present invention is not limited by the specific implementation modes.

[0035] Unless explicitly stated otherwise, throughout the specification and claims, the term “comprise” or variations such as “include” or “comprising”, etc., will be understood to include the stated elements or components but not to exclude other elements or components.

[0036] In this document, for the convenience of description, spatial relative terms such as "below", "below", "down", "above", "above", "upper", etc. may be used to describe the relationship between one element or feature and another element or feature in the accompanying drawings. It should be understood that the spatial relative terms are intended to include different orientations of the object in use or operation in addition to the orientation depicted in the figure. For example, if the object in the figure is turned over, the element described as being "below" or "below" other elements or features will be oriented "above" the element or feature. Therefore, the exemplary term "below" can include both below and above. Objects may also have other orientations (rotated 90 degrees or other orientations) and the spatial relative terms used in this document should be interpreted accordingly.

[0037] In this document, the terms "first", "second", etc. are used to distinguish two different elements or parts, and are not used to limit a specific position or relative relationship. In other words, in some embodiments, the terms "first", "second", etc. can also be interchangeable.

[0038] Embodiment 1

[0039] In order to obtain a more accurate oil consumption forecasting model, and provide effective data support for the distribution of refined oil products and the establishment of refined oil supply chains, such as Figure 1 As shown, in an embodiment of the present invention, a method for predicting the sales volume of refined oil at a gas station is provided, comprising the steps of:

[0040] S11, obtaining liquid level data of a liquid level meter at a gas station, and generating a first array based on a time series; the elements of the first array include liquid level data and corresponding time data;

[0041] In the prior art, sales data is generally used as modeling data. The inventors have found through research that the reasons why the prediction models in the prior art have poor accuracy include at least the following two aspects:

[0042] First, there is a large time lag between the generation of sales data and the change of fuel in gas stations, resulting in a poor correspondence between the time point of sales data and the actual sales time point; on the other hand, the amount of sales data is very small and has poor periodicity.

[0043] Based on the above understanding, in an embodiment of the present invention, sales data is no longer used as modeling data, but liquid level data is used as raw data and processed into modeling data.

[0044] Liquid level data can be collected by a liquid level meter at a gas station. In actual applications, the data collection cycle can be one data every three seconds, so that raw data with very small granularity and large data volume can be obtained.

[0045] S12, sampling the elements of the first array according to a preset time period to generate a second array;

[0046] The elements of the first array include a large amount of liquid level data and corresponding time data. Through periodic sampling, the amount of modeling data can be reduced to a level suitable for building a prediction model. In the embodiment of the present invention, the time period is preferably 30 minutes.

[0047] In practical applications, the sampling period may also be determined by those skilled in the art based on experience or a limited number of experiments, and is not specifically limited here.

[0048] S13, generating a third array according to the second array, comprising: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array respectively to obtain a plurality of oil consumption data; establishing a correspondence between the oil consumption data and the time data in the rear element; generating the third array according to the correspondence between the oil consumption data and the time data;

[0049] The simple liquid level data cannot directly represent the consumption of oil products. Therefore, in the embodiment of the present invention, the third array is generated according to the second array, that is, the oil product consumption data is generated by the original liquid level data. The specific method may be:

[0050] Subtract the liquid level data in the front and back adjacent elements in the second array to obtain multiple oil consumption data; since the second array includes the liquid level data at periodic time points, the corresponding oil consumption data can be obtained by subtracting the rear liquid level data from the front liquid level data, and so on, the third data can be obtained to characterize the corresponding relationship between each oil consumption data and time.

[0051] S14, determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data;

[0052] Since gas stations will refill oil regularly or irregularly, at this time, the latter liquid level data in the second data will be greater than the former liquid level data, which will cause the oil consumption data in the second data to be negative. Therefore, the third array cannot be directly used for modeling.

[0053] To correct abnormal data, the average of the previous oil product consumption data and the next oil product consumption data can be used to replace the abnormal data. It should be noted that the previous oil product consumption data and the next oil product consumption data should both be normal data.

[0054] In the embodiment of the present invention, the abnormal data may be modified specifically according to the characteristics of the gas station, specifically:

[0055] The inventor has found through research that the oil consumption data of a gas station has a general fluctuation period, during which the oil consumption data at the same time point are basically similar; in practical applications, the fluctuation period can be set to one week.

[0056] Based on the above findings, in the embodiment of the present invention, the time point at which the abnormal data is located in the fluctuation period is used as the corresponding data to determine the oil consumption data at the time point in the multiple fluctuation periods before and after it as the reference data, and then correct the abnormal data according to the multiple reference data, specifically:

[0057] The data correction includes: obtaining a plurality of reference data corresponding to the abnormal data in a third array; calculating the mean of the plurality of reference data to replace the abnormal data; the reference data is the oil consumption data with a fluctuation period of one week and at the same day and time as the abnormal data.

[0058] In addition, since the abnormal data is caused by the gas station refueling the gas station, in the embodiment of the present invention, the abnormal data can also be corrected according to the fuel replenishment data, specifically:

[0059] Real-time collection of fuel replenishment data for each fuel replenishment;

[0060] When data anomaly occurs, obtain the fuel replenishment data closest to the time data corresponding to the abnormal data;

[0061] The abnormal data is corrected according to the oil replenishment data, that is, the negative oil consumption data is added to the oil replenishment data to obtain the actual oil consumption data.

[0062] S15. Using the elements of the third array as modeling data, an oil consumption prediction model is generated through data training, wherein the oil consumption prediction model is used to predict oil consumption prediction data for a preset number of days in the future.

[0063] In practical applications, the oil sales volume of a gas station will also be affected by factors such as the number of gas pumps, geographical location, holiday data, and weather information data. Therefore, the data of these influencing factors can also be used as modeling data together with the elements of the third array.

[0064] Preferably, the embodiment of the present invention can be based on a generative adversarial prediction method, and the oil consumption prediction model can output single-step prediction information through multi-step input data to achieve single-step prediction, and can also output multi-step prediction information through multi-step input data to achieve multi-step prediction. The oil consumption prediction model can include a non-recursive multi-step prediction model, wherein the multi-step prediction means using the oil consumption data of the past preset number of days to predict the oil consumption data of the future preset number of days.

[0065] In an embodiment of the present invention, an oil consumption prediction model can be generated by two long short-term memory networks LSTM; specifically, the two LSTMs are respectively connected to the same neural network to respectively realize single-step prediction and multi-step prediction; the neural network is used to predict changes in data at future moments by learning the change rules in continuous data.

[0066] Furthermore, in the embodiment of the present invention, the following may also be included:

[0067] S16. Determine a fuel replenishment plan based on the forecast data of fuel consumption for a preset number of days in the future and the current inventory of refined fuel products at the gas station.

[0068] After the oil consumption prediction model in the embodiment of the present invention has predicted the future oil consumption, the date and amount of oil replenishment can be reasonably planned and arranged.

[0069] In summary, the embodiment of the present invention no longer uses the sales data of the gas station as the modeling data for constructing the prediction model, but collects and stores the liquid level data of the gas station liquid level meter in real time to generate modeling data with better data granularity and timeliness; when generating the modeling data, the embodiment of the present invention obtains the fuel consumption data at each time point by calculating the liquid level difference between the adjacent data; in this way, after constructing the modeling data consisting of the fuel consumption data and the corresponding time; it is possible to generate an fuel consumption prediction model for predicting the fuel consumption prediction data for a preset number of days in the future through data training.

[0070] It can be seen from the above that, compared with the sales data used in the prior art, the change time points of the oil data in the modeling data in the embodiment of the present invention correspond better to the actual sales time points of the oil; in addition, in the embodiment of the present invention, the collection period of the liquid level data of the gas station level meter is very short, so the amount of data generated is very large, so that sufficient modeling data can be generated according to different sampling periods; that is, the modeling data in the embodiment of the present invention has stronger timeliness and a larger amount of data, so it is possible to construct an oil consumption prediction model with better prediction accuracy; this can also provide effective data support for the distribution of refined oil and the establishment of a refined oil supply chain.

[0071] Embodiment 2

[0072] In another aspect of the embodiment of the present invention, a device for predicting the sales volume of refined oil at a gas station is also provided. Figure 2 The schematic diagram of the structure of the device for predicting the sales volume of refined oil at a gas station provided by an embodiment of the present invention is shown. Figure 1 The device corresponding to the method for predicting the sales volume of refined oil at a gas station described in the corresponding embodiment, that is, the virtual device is used to implement Figure 1 In the corresponding embodiment of the method for predicting the sales volume of refined oil at a gas station, each virtual module constituting the device for predicting the sales volume of refined oil at a gas station can be executed by an electronic device, such as a network device, a terminal device, or a server. Specifically, the device for predicting the sales volume of refined oil at a gas station in the embodiment of the present invention includes:

[0073] The data acquisition unit 01 is used to acquire the liquid level data of the liquid level meter at the gas station and generate a first array based on the time series; the elements of the first array include the liquid level data and the corresponding time data;

[0074] A data sampling unit 02, configured to sample the elements of the first array according to a preset time period to generate a second array;

[0075] The modeling data generating unit 03 is used to generate a third array according to the second array, including: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array to obtain a plurality of oil consumption data; establishing a corresponding relationship between the oil consumption data and the time data in the rear element; and generating the third array according to the corresponding relationship between the oil consumption data and the time data;

[0076] A modeling data correction unit 04 is used to determine the oil consumption data with negative values ​​in the third array as abnormal data, and perform data correction on the abnormal data;

[0077] The model building unit 05 is used to use the elements of the third array as modeling data to generate an oil consumption prediction model through data training, and the oil consumption prediction model is used to predict the oil consumption prediction data of a preset number of days in the future.

[0078] Preferably, the present invention further comprises:

[0079] The plan generation unit (not shown in the figure) is used to determine the oil replenishment plan based on the oil consumption forecast data for a preset number of days in the future and the current inventory of refined oil products at the gas station.

[0080] Since the working principle and beneficial effects of the device for predicting the sales volume of refined oil at a gas station in the embodiment of the present invention have been Figure 1 The corresponding method for predicting the sales volume of refined oil at gas stations is also recorded and explained, so they can be cross-referenced and will not be repeated here.

[0081] Embodiment 3

[0082] Corresponding to the method embodiment, in the embodiment of the present invention, a device for predicting the sales volume of refined oil at a gas station is also provided, such as a terminal, a server, etc. Among them, the server can be an independent physical server, or a server cluster or a distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The terminal can be a smart phone, a tablet computer, a laptop computer, a desktop computer, etc., but is not limited thereto.

[0083] An example diagram of a hardware structure block diagram of a gas station refined oil sales volume prediction device provided by an embodiment of the present invention is shown in FIG. Figure 3 As shown, it may include:

[0084] Processor 1, communication interface 2, memory 3 and communication bus 4;

[0085] The processor 1, the communication interface 2, and the memory 3 communicate with each other via the communication bus 4;

[0086] Optionally, the communication interface 2 may be an interface of a communication module, such as an interface of a GSM module;

[0087] The processor 1 may be a central processing unit CPU, or an application-specific integrated circuit ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0088] The memory 3 may include a high-speed RAM memory, and may also include a non-volatile memory (non-volatile memory), such as at least one disk memory.

[0089] The processor 1 is specifically used to execute the computer program stored in the memory 3 to perform the following steps:

[0090] S11, obtaining liquid level data of a liquid level meter at a gas station, and generating a first array based on a time series; the elements of the first array include liquid level data and corresponding time data;

[0091] S12, generating a second array by sampling elements of the first array according to a preset time period;

[0092] S13, generating a third array according to the second array, comprising: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array respectively to obtain a plurality of oil consumption data; establishing a correspondence between the oil consumption data and the time data in the rear element; generating the third array according to the correspondence between the oil consumption data and the time data;

[0093] S14, determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data;

[0094] S15. The elements of the third array are modeling data. An oil consumption prediction model is generated through data training. The oil consumption prediction model is used to predict oil consumption prediction data for a preset number of days in the future.

[0095] Preferably, the present invention further comprises:

[0096] S16. Determine a fuel replenishment plan based on the fuel consumption forecast data for a preset number of days in the future and the current inventory of refined oil products at the gas station.

[0097] The above product can execute the method provided in the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not fully described in this embodiment, please refer to the method for predicting the sales volume of refined oil products at a gas station provided in the embodiment of the present invention.

[0098] Embodiment 4

[0099] In an embodiment of the present invention, a storage medium is further provided, which may store a program suitable for execution by a processor, wherein the program is used to:

[0100] S11, obtaining liquid level data of a liquid level meter at a gas station, and generating a first array based on a time series; the elements of the first array include liquid level data and corresponding time data;

[0101] S12, generating a second array by sampling elements of the first array according to a preset time period;

[0102] S13, generating a third array according to the second array, comprising: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array respectively to obtain a plurality of oil consumption data; establishing a correspondence between the oil consumption data and the time data in the rear element; generating the third array according to the correspondence between the oil consumption data and the time data;

[0103] S14, determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data;

[0104] S15. The elements of the third array are modeling data. An oil consumption prediction model is generated through data training. The oil consumption prediction model is used to predict oil consumption prediction data for a preset number of days in the future.

[0105] Preferably, the present invention further comprises:

[0106] S16. Determine a fuel replenishment plan based on the fuel consumption forecast data for a preset number of days in the future and the current inventory of refined oil products at the gas station.

[0107] Optionally, the detailed functions and extended functions of the program may refer to the above description.

[0108] The above product can execute the method provided by the embodiment of the present invention, and has the functional modules and beneficial effects corresponding to the execution method. For technical details not described in detail in this embodiment, please refer to the methods provided by other embodiments of the present invention.

[0109] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0110] In the several embodiments provided in this application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or unit, which can be electrical, mechanical or other forms.

[0111] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0112] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0113] It should be understood that in the embodiments of the present application, the various embodiments and features can be combined with each other to solve the aforementioned technical problems.

[0114] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0115] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the sales volume of refined oil at a gas station, characterized in that: Includes steps: S11, obtaining liquid level data of a liquid level meter at a gas station, and generating a first array based on a time series; the elements of the first array include liquid level data and corresponding time data; S12, sampling the elements of the first array according to a preset time period to generate a second array; S13, generating a third array according to the second array, comprising: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array respectively to obtain a plurality of oil consumption data; establishing a correspondence between the oil consumption data and the time data in the rear element; generating the third array according to the correspondence between the oil consumption data and the time data; S14, determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data; S15. Using the elements of the third array as modeling data, an oil consumption prediction model is generated through data training, wherein the oil consumption prediction model is used to predict oil consumption prediction data for a preset number of days in the future.

2. The method for predicting the sales volume of refined oil at a gas station according to claim 1, characterized in that: Also includes: S16. Determine a fuel replenishment plan based on the fuel consumption forecast data for a preset number of days in the future and the current inventory of refined oil products at the gas station.

3. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The data correction includes: The abnormal data is replaced by the average of the previous oil product consumption data and the next oil product consumption data of the abnormal data.

4. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The data correction includes: A plurality of reference data corresponding to the abnormal data are obtained in the third array; an average of the plurality of reference data is calculated to replace the abnormal data; the reference data is: oil consumption data at the same time and on the same day as the abnormal data with a fluctuation period of one week.

5. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The data correction includes: Acquire the fuel replenishment data closest to the time data corresponding to the abnormal data; The abnormal data is corrected according to the fuel replenishment data.

6. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The time period includes: The time period is 30 minutes.

7. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The modeling data also includes: One or more of the number of gas pumps, geographic location, holiday data, and weather information data.

8. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The oil consumption prediction model includes a non-recursive multi-step prediction model, wherein the multi-step prediction means using the oil consumption data of the past preset number of days to predict the oil consumption data of the future preset number of days.

9. The method for predicting the sales volume of refined oil at a gas station according to claim 1 or 2, characterized in that: The oil consumption prediction model is generated by two long short-term memory networks (LSTM), including: The two LSTMs are respectively connected to the same neural network to realize single-step prediction and multi-step prediction respectively; the neural network is used to predict the changes of data at future moments by learning the change rules in continuous data.

10. A device for predicting the sales volume of refined oil at a gas station, characterized in that: include: A data acquisition unit, used to acquire liquid level data of a liquid level meter at a gas station, and generate a first array based on a time series; the elements of the first array include liquid level data and corresponding time data; A data sampling unit, used for sampling the elements of the first array according to a preset time period to generate a second array; The modeling data generating unit is used to generate a third array according to the second array, including: performing subtraction operations on the liquid level data in the front and rear adjacent elements in the second array to obtain a plurality of oil consumption data; establishing a corresponding relationship between the oil consumption data and the time data in the rear element; generating the third array according to the corresponding relationship between the oil consumption data and the time data; A modeling data correction unit, used for determining the oil consumption data with negative values ​​in the third array as abnormal data, and performing data correction on the abnormal data; The model building unit is used to use the elements of the third array as modeling data to generate an oil consumption prediction model through data training, wherein the oil consumption prediction model is used to predict the oil consumption prediction data for a preset number of days in the future.

11. The device for predicting the sales volume of refined oil at a gas station according to claim 10, characterized in that: Also includes: The plan generation unit is used to determine the oil replenishment plan based on the oil consumption forecast data for a preset number of days in the future and the current inventory of refined oil at the gas station.

12. A device for predicting the sales volume of refined oil at a gas station, characterized in that: include: Memory for storing computer programs; A processor is used to call and execute the computer program to implement the steps of the method for predicting the sales volume of refined oil at a gas station as described in any one of claims 1-7.

13. A storage medium, characterized in that: It comprises a software program, which is suitable for a processor to execute the steps of the method for predicting the sales volume of finished oil at a gas station as described in any one of claims 1-7.

Images