Building heating load prediction method based on artificial intelligence algorithm

Abstract

The invention belongs to the technical field of heat supply, and particularly relates to a building heat supply load prediction method based on an artificial intelligence algorithm. Comprising the following steps: S1, acquiring weather data and outdoor temperature data; s2, acquiring heat supply demand indoor temperature data of a user; s3, obtaining a user heat supply unit area heat load index according to the outdoor temperature data and the heat supply demand indoor temperature data; s4, according to the heat supply load corresponding to the historical similar weather data corresponding to the weather data and the heat supply effect data of the historical similar weather days; s5, comparing the heat supply data required by the user with historical similar weather day heat supply data to obtain a first correction coefficient; s6, obtaining a deviation auxiliary correction coefficient of the first correction coefficient according to the weather data; and S7, according to the user heat supply unit area heat load index, the heat supply load corresponding to the historical similar weather data, the first correction coefficient and the deviation auxiliary correction coefficient, obtaining building heat supply load prediction data. According to the method, prediction is intelligent, resources are saved, and the requirements of customers can be met.

Description

Technical Field

[0001] The present invention relates to the technical field of heating, and particularly relates to a method for predicting building heating load based on an artificial intelligence algorithm. Background Art

[0002] The description of the background art in the present invention belongs to the related art related to the present invention, and is only used to illustrate and facilitate the understanding of the content of the present invention, and should not be construed as the applicant clearly believing or presuming that the applicant believes it is the prior art on the filing date of the first application of the present invention.

[0003] With the development of society and the improvement of people's living standards, people's requirements for heating are also getting higher and higher. However, at present, heating generally adopts a rough heating method, without a clear prediction of the heating load, and uses manual control methods. For example, in some areas, it completely depends on the boiler operator to roughly judge according to the weather or control the heating according to the energy budget. Such a rough method has many defects. For example, in warm weather, it not only causes waste of resources but also the temperature is too high and the user experience is not good. Moreover, in colder weather, the required room temperature cannot be reached, resulting in a very poor user experience, which intensifies the contradiction between users and heating enterprises.

[0004] Generally, for residential buildings: the general heating load index is 20 - 30 kWh / m² / year. This range includes variations under different regions and climate conditions. For example, in cold regions or northern cities, due to the lower winter temperatures, more energy is required for heating, so the heating load index may be higher.

[0005] For commercial and office buildings: their heating load index is usually slightly higher than that of residential buildings. According to the differences in specific functions and usage requirements, the heating demand of commercial and office buildings may be between 30 - 50 kWh / m² / year.

[0006] For new buildings: with the increasing requirements for energy conservation and environmental protection and the application of new technologies, more and more new buildings adopt high - efficiency energy - saving technologies and meet the heating demand by comprehensively using renewable energy such as solar energy and ground - source heat pumps. The heating load index of these new buildings may be lower, usually between 10 - 20 kWh / m² / year. It should be noted that these values are for reference only, and the actual heating load index needs to be calculated and designed in detail according to the specific building conditions and local climate conditions.

[0007] Some literature has also studied the heating load. "Civil Heating Radiators" (Tsinghua University Press) mentions that the influencing factors of heating load include heating area, heat index per unit area, and correction coefficient. However, the correction coefficient described in it is relatively single. Different environments and different demands will surely have different impacts on the heating load. There is no single coefficient that can meet the loads caused by different demands. This is undoubtedly the case. Therefore, there is an urgent need for an intelligent heating load prediction method to solve this problem. Summary of the Invention

[0008] The purpose of the embodiments of the present invention is to provide a building heating load prediction method based on artificial intelligence algorithms. The method of the present invention combines historical data, customer demand data, weather data, and correction coefficients for the impact of different weathers on heating demand, making the heating prediction more intelligent. While saving resources, it can also meet the needs of customers.

[0009] The purpose of the present invention is achieved by the following method:

[0010] A building heating load prediction method based on artificial intelligence algorithms includes the following steps:

[0011] S1. Obtain weather data and outdoor temperature data;

[0012] S2. Obtain the indoor temperature data of the heating demand of the user;

[0013] S3. Obtain the heat load index per unit area of the user's heating according to the outdoor temperature data and the indoor temperature data of the heating demand;

[0014] S4. According to the heating load corresponding to the historical similar weather data corresponding to the weather data and the heating effect data of the historical similar weather days;

[0015] S5. Compare the heating data required by the user with the heating data of the historical similar weather days to obtain the first correction coefficient;

[0016] S6. Obtain the deviation auxiliary correction coefficient of the first correction coefficient according to the weather data;

[0017] S7. Obtain the building heating load prediction data according to the heat load index per unit area of the user's heating, the heating load corresponding to the historical similar weather data, the first correction coefficient, and the deviation auxiliary correction coefficient.

[0018] Furthermore, the calculation method of step S7 is as follows:

[0019]

[0020] Wherein, M 预测The predicted value of the building heating load; K1 is the coefficient of the heat load per unit area of the user's heating; q is the heat load index per unit area; S is the heating area; K2 is the heating load coefficient corresponding to the historical similar weather data; Q is the heating load corresponding to the historical similar weather data; L1 is the first correction coefficient; L2 is the deviation auxiliary correction coefficient.

[0021] Further, the values of K1 and K2 are determined in the following manner:

[0022] K1 + K2 = 1;

[0023] If the absolute value of F - F0 is less than or equal to 0.3 times of F0, then

[0024] If the absolute value of F - F0 is greater than 0.3 times of F0, then K2 = 0.01;

[0025] Wherein, F is the historical similar weather data; F0 is the weather data.

[0026] Further, the algorithm of F - F0 is as follows;

[0027] F - F0 = ∑N(D1 - D0) i

[0028] Wherein, D0 is the data after digital processing of an index in the weather data; D1 is the data after digital processing of an index in the historical similar weather data; N is the weight corresponding to each index.

[0029] Further, L2 is determined in the following manner: Define the weather factors that cause high humidity as positive values, and the weather factors that do not cause high humidity as 0.

[0030] Further, the user can set the temperature requirements according to time periods, calculate the heating loads according to different requirements for different time periods, and obtain the total heating load.

[0031] The embodiments of the present invention have the following beneficial effects:

[0032] The method of the present invention combines historical data, customer demand data, weather data, and the influence correction coefficient of different weathers on heating demand, making the prediction of heating more intelligent. While saving resources, it can also meet the customer's needs. Specific implementation manners

[0033] The following further introduces the present application in combination with embodiments.

[0034] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, in the following description, different "one embodiment" or "embodiment" do not necessarily refer to the same embodiment. Different embodiments can be replaced or combined. For those of ordinary skill in the art, other implementation manners can be obtained based on these embodiments without creative efforts.

[0035] A building heating load prediction method based on an artificial intelligence algorithm includes the following steps:

[0036] S1. Obtain weather data and outdoor temperature data;

[0037] S2. Obtain the indoor temperature data of the heating demand of users;

[0038] S3. Obtain the heat load index per unit area of user heating according to the outdoor temperature data and the indoor temperature data of the heating demand;

[0039] S4. According to the heating load corresponding to the historical similar weather data corresponding to the weather data and the heating effect data of the historical similar weather days;

[0040] S5. Compare the required heating data of the user with the heating data of the historical similar weather days to obtain the first correction coefficient;

[0041] S6. Obtain the deviation auxiliary correction coefficient of the first correction coefficient according to the weather data;

[0042] S7. Obtain the building heating load prediction data according to the heat load index per unit area of user heating, the heating load corresponding to the historical similar weather data, the first correction coefficient, and the deviation auxiliary correction coefficient.

[0043] The method of the present invention combines historical data, customer demand data, weather data, and the influence correction coefficient of different weathers on heating demand, making the prediction of heating more intelligent. While saving resources, it can also meet the needs of customers.

[0044] In some embodiments, the calculation method of step S7 is as follows:

[0045]

[0046] Among them, M 预测 is the predicted value of the building heating load; K1 is the heat load index coefficient per unit area of user heating; q is the heat load index per unit area; S is the heating area; K2 is the heating load coefficient corresponding to the historical similar weather data; Q is the heating load corresponding to the historical similar weather data; L1 is the first correction coefficient; L2 is the deviation auxiliary correction coefficient.

[0047] The calculation method in the above embodiments is explained as follows: The product of qS can calculate the heating load obtained from indoor and outdoor temperature data, but this is only a rough and simple calculation method, and it is also the general data range for determining the heating load. It is allocated with different coefficients from the actual heating load generated under similar historical weather conditions to achieve the heating load jointly determined by the actual heating load in the past and the heating load predicted by the existing weather data. However, these two data are obviously not accurate and still need some corrections. Although K2 is calculated based on the deviation between similar historical weather data and weather data (the specific calculation method will be described later), the applicant found that just making such corrections, there is still a certain gap between the calculated heating load and the actual heating load. After a large amount of research and practice, the applicant determined to correct the coefficient of the heating load on similar historical weather days. L1 is the absolute value of the ratio of the weather data of similar historical weather days to the weather data of the current day. The problem here is that if the correction is simply calculated in this way, the effect is not ideal. The applicant also corrected the first correction coefficient, and the determination method of this auxiliary correction will be described in detail later and will not be elaborated here.

[0048] In some embodiments of the present invention, the values of K1 and K2 are determined in the following manner:

[0049] K1 + K2 = 1;

[0050] If the absolute value of F - F0 is less than or equal to 0.3 times of F0, then

[0051] If the absolute value of F - F0 is greater than 0.3 times of F0, then K2 = 0.01;

[0052] Wherein, F is similar historical weather data; F0 is weather data.

[0053] Furthermore, the algorithm of F - F0 is as follows;

[0054] F - F0 = ∑N(D1 - D0) i

[0055] Wherein, D0 is the data after digital processing of an index in the weather data; D1 is the data after digital processing of an index in the similar historical weather data; N is the weight corresponding to each index.

[0056] For the convenience of understanding the purpose of the design of this calculation method, the following detailed description is provided: The weather data contains multiple data, such as sunny or cloudy, rain or snow, ultraviolet rays, light intensity, and wind force, etc. Here, each heating enterprise can determine according to its own situation and the actual heating situation, as long as the sum of each weight conforms to common sense and the weight items are important factors affecting heating. Different regions are different. For example, in the Northeast, there is more snow accumulation, which will obviously affect the heating effect, and its weight will necessarily be larger. While in North China, there is less snow in winter and it will not accumulate for a long time, so the impact is smaller and its weight can be appropriately reduced. The applicant will not elaborate on this one by one here.

[0057] In some embodiments, L2 is determined in the following manner: The weather factors that cause high humidity are defined as positive values, and the weather factors that do not cause high humidity are defined as 0. It can be understood that: High humidity will affect the perceived temperature. For example, the heating temperature is 22 degrees, but the indoor perceived temperature will be colder in humid and cold weather, so the heating should be appropriately increased, and the heating load will also increase. While in sunny days with good sunlight, the indoor perceived temperature is comfortable, so this corrected part can be set to 0.

[0058] In other embodiments, the user can set the temperature requirements according to time periods, calculate the heating load according to different requirements for different time periods, and then obtain the total heating load. It can be understood that different family conditions, different heat demands and other factors of each family all result in different demands. In the past, the heating was all with unified heating parameters and the users could not adjust them, resulting in extremely poor experience. This application takes this into account. For example, when the user goes out to work during the day and there is no one at home for most of the time, the heating can be automatically reduced appropriately. While at night when sleeping, the heating requirement is higher, so the temperature can be set to increase appropriately. It should be understood that the premise of adopting this method should be that the heating is charged according to the amount, which is beneficial for the customers to set according to their own needs, can also reduce the operation cost while saving resources, and does not reduce the heating experience. It is intelligent and adopts different heating methods for different users at different time periods according to their needs, without causing waste of resources.

[0059] It should be noted that the above embodiments can be freely combined according to needs. The above introduction is only the preferred embodiments of the present invention and is not used to limit the present invention. For those skilled in the art, the present invention can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for predicting building heating load based on artificial intelligence algorithm, characterized in that, It includes the following steps: S1. Obtain weather data and outdoor temperature data; S2. Obtain the indoor temperature data of the user's heating demand; S3. Obtain the heat load index per unit area of the user's heating according to the outdoor temperature data and the indoor temperature data of the heating demand; S4. According to the heating load corresponding to the historical similar weather data corresponding to the weather data and the heating effect data of the historical similar weather days; S5. Compare the heating data required by the user with the heating data of the historical similar weather days to obtain the first correction coefficient; S6. Obtain the deviation auxiliary correction coefficient of the first correction coefficient according to the weather data; S7. Obtain the predicted building heating load data according to the heat load index per unit area of the user's heating, the heating load corresponding to the historical similar weather data, the first correction coefficient and the deviation auxiliary correction coefficient.

2. The building heating load prediction method based on artificial intelligence algorithm according to claim 1, wherein The calculation method of step S7 is as follows: Among them, M 预测 is the predicted value of the building heating load; K1 is the heat load index coefficient per unit area of the user's heating; q is the heat load index per unit area; S is the heating area; K2 is the heating load coefficient corresponding to the historical similar weather data; Q is the heating load corresponding to the historical similar weather data; L1 is the first correction coefficient; L2 is the deviation auxiliary correction coefficient.

3. The building heating load prediction method based on artificial intelligence algorithm according to claim 2, characterized in that, The values of K1 and K2 are determined in the following way: K1 + K2 = 1; If the absolute value of F - F0 is less than or equal to 0.3 times F0, then If the absolute value of F - F0 is greater than 0.3 times of F0, then K2 = 0.01; Wherein, F is the historical similar weather data; F0 is the weather data.

4. The method for predicting building heating load based on artificial intelligence algorithm according to claim 3, wherein The algorithm of F - F0 is as follows; F - F0 = ∑N(D1 - D0) i Wherein, D0 is the data after digital processing of an index in the weather data; D1 is the data after digital processing of an index in the historical similar weather data; N is the weight corresponding to each index.

5. The building heating load prediction method based on an artificial intelligence algorithm according to claim 3, wherein The determination of L2 is determined in the following way: Define the weather factors causing high humidity as positive values, and define the weather factors not causing high humidity as 0.

6. The method for predicting building heating load based on artificial intelligence algorithm according to claim 3, wherein The user can set the temperature demand according to the time period, calculate the heating load according to different demands in different time periods, and then obtain the total heating load.

Images