Sunlight prediction method and device, computer equipment and storage medium

Abstract

The invention relates to the technical field of sunlight analysis, and discloses a sunlight prediction method and device, computer equipment and a storage medium, and the method comprises the steps: employing random number parameterization to generate a training data set, and determining a target sunlight prediction model; obtaining geographic information of a target area, geometric attribute information of each building in the target area and three-dimensional coordinates of a target space point, performing geometric coordinate transformation, determining a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point, and inputting each parameter into a target sunlight prediction model to obtain a target sunlight prediction model; and determining the total duration of shielding the target space point by each building and the shielding starting moment, further calculating to obtain the total duration of shielding the target space point, and finally determining the sunshine duration of the target space point. The target sunlight prediction model based on the neural network is adopted, the calculation efficiency is improved, a fixed training data set is not needed, the storage space is saved, and the convergence difficulty and the fitting error of the neural network are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of sunshine analysis, and in particular to a sunshine prediction method, device, computer equipment and storage medium. Background Art

[0002] Natural sunlight not only affects the heating, cooling, lighting and energy consumption of buildings, but also affects the health of residents. Usually, designers use the center of the ground floor window as the calculation height for planar sunlight analysis, and calculate the direct sunlight duration of each point on the plane at that height. If the spatial point to the south of the projection of the building on the calculation plane and closest to the projection meets the sunlight requirements, it can be inferred that the building meets the sunlight requirements.

[0003] There are currently a variety of sunlight analysis software on the market to ensure that building designs comply with national standards. Usually, when these software are performing sunlight analysis, if the analysis time step is increased, the number of shadow analyses will decrease, affecting the calculation accuracy; if the number of analyses is increased, the calculation time will be too long, and it is difficult to ensure high accuracy and low calculation time at the same time, and the calculation efficiency is not high. Summary of the invention

[0004] In view of this, an embodiment of the present invention provides a sunshine prediction method, apparatus, computer equipment and storage medium to solve the problem of low sunshine calculation efficiency.

[0005] In a first aspect, an embodiment of the present invention provides a sunshine prediction method, the method comprising:

[0006] Generate training data set based on random number parameterization and determine target sunshine prediction model;

[0007] Obtaining geographic information of the target area, geometric attribute information of each building in the target area, and three-dimensional coordinates of the target space point;

[0008] Performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point;

[0009] For each of the buildings, the first geometric parameter, the second coordinate parameter and the geographic information of the target area are respectively input into a target sunshine prediction model to determine the total duration and the start time of the shielding of the target spatial point by each of the buildings;

[0010] Based on the total duration of shielding of the target space point by each building and the shielding start time, the total duration of shielding of the target space point is obtained;

[0011] The sunshine duration of the target space point is determined based on the total sunshine duration of the target area and the total duration of shielding of the target space point.

[0012] In some optional implementations, obtaining the total duration of shielding of the target space point based on the total duration of shielding of the target space point by each building and the shielding start time includes:

[0013] Calculate the shading end time of the target space point based on the total duration of the target space point being shaded by each building and the shading start time;

[0014] Based on the shielding start time and the shielding end time, the total duration of shielding of the target space point is obtained.

[0015] In some optional implementations, obtaining the total duration of shielding of the target space point based on the shielding start time and the shielding end time includes:

[0016] The shielding start time and the shielding end time are sorted respectively to determine the shielding start time sequence and the shielding end time sequence of the spatial point;

[0017] Calculate the difference between the maximum time in the masking end time sequence and the minimum time in the masking start time sequence to determine a first total duration;

[0018] The unobstructed time period between the maximum moment in the obstruction end time sequence and the minimum moment in the obstruction start time sequence is calculated, and the difference between the first total duration and the unobstructed time period is calculated to obtain the total duration of the target space point being obscured.

[0019] In some optional implementations, the geometric attribute information of the building includes building dimensions, the building dimensions include length, width and height, and performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point includes:

[0020] Determine the reference size by subtracting the height of the target space point from the sum of the length, width and height of the building, where the height of the target space point is determined by the three-dimensional coordinates of the target space point;

[0021] Determine the relative length, relative width and relative height of the building based on the reference size, the three-dimensional coordinates of the target space point and the building size;

[0022] Determining the orientation of the building based on the angle between the building size and the preset orientation, wherein the first geometric parameter of the building includes a relative length, a relative width, a relative height and an orientation;

[0023] A second coordinate parameter of the target space point is determined based on a ratio of the three-dimensional coordinate of the target space point to the reference size, wherein the second coordinate parameter includes a two-dimensional polar coordinate.

[0024] In some optional implementations, the generating a training data set based on random number parameterization and determining a target sunshine prediction model includes:

[0025] Generate features of the training data set based on random number parameterization, the features including a first geometric parameter, a second coordinate parameter, and geographic information of a training area;

[0026] Input the features of the training data set into a reference sunshine prediction model, and output the true value of the training data set label;

[0027] Input the features of the training data set into a target sunshine prediction model, and output a predicted value of the training data set label;

[0028] The loss is calculated according to the true value and the predicted value of the label, and the parameters of the target sunshine prediction model are updated based on the loss to determine the target sunshine prediction model.

[0029] In some optional implementations, the feature of generating the training data set based on random number parameterization includes:

[0030] Generate geographic information of the training area, geometric information of virtual buildings in the training area, and heights of virtual space points based on random number parameterization;

[0031] Determine a first geometric parameter of the virtual building based on the geometric information of the virtual building in the training area and the height of the virtual space point;

[0032] Based on a preset first step length sampling network and a second step length sampling network, a second coordinate parameter of the virtual space point is determined in a preset sampling area, and the first step length is greater than the second step length.

[0033] In a second aspect, an embodiment of the present invention provides a sunshine prediction device, the device comprising:

[0034] A data acquisition module is used to acquire the geographic information of the target area, the geometric attribute information of each building in the target area, and the three-dimensional coordinates of the target space point;

[0035] A data processing module, used for performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point, and determining a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point;

[0036] A parameter calculation module, for inputting the first geometric parameter, the second coordinate parameter and the geographic information of the target area into a target sunshine prediction model for each of the buildings, and determining a total duration and a start time of shielding of the target spatial point by each of the buildings;

[0037] A first calculation module is used to obtain the total time duration of the target space point being shielded based on the total time duration of the target space point being shielded by each building and the shielding start time;

[0038] The second calculation module is used to determine the sunshine duration of the target space point based on the total sunshine duration of the target area and the total duration of the target space point being shaded.

[0039] In a third aspect, an embodiment of the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, computer instructions being stored in the memory, and the processor executing the sunshine prediction method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0040] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the sunshine prediction method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 is a schematic flow chart of a sunshine prediction method according to some embodiments of the present invention;

[0043] Figure 2 is a schematic diagram of a parameter calculation method according to some embodiments of the present invention;

[0044] Figure 3 is a flow chart of another sunshine prediction method according to some embodiments of the present invention;

[0045] Figure 4 is a schematic diagram of a parameter calculation method according to some embodiments of the present invention;

[0046] Figure 5 is a flow chart of another sunshine prediction method according to some embodiments of the present invention;

[0047] Figure 6 is a schematic flow chart of a mixed step-size floating starting point calculation point sampling method according to some embodiments of the present invention;

[0048] Figure 7 is a schematic diagram of a target sunshine prediction model according to some embodiments of the present invention;

[0049] Figure 8 is a structural block diagram of a sunshine prediction device according to an embodiment of the present invention;

[0050] Fig. 9 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0052] Natural sunlight affects the heating, cooling, lighting and energy consumption of buildings, and also affects the health of residents in residential buildings. According to regulations, all windows on the south facade of newly built residential buildings must meet the minimum number of direct sunlight hours within a specific time period. For example, cities with a permanent population of more than 500,000 in cold regions must ensure more than 2 hours of sunlight within 8 to 16 hours on the Great Cold Day, and cities with a permanent population of less than 500,000 in hot summer and warm winter regions must ensure more than 1 hour of sunlight within 9 to 15 hours on the Winter Solstice.

[0053] Architects will use sunshine analysis software to simulate the sunshine thermal field diagram of the design area plane during the architectural design stage, and optimize the design plan according to the simulation results. Based on this, an embodiment of the present invention provides a sunshine prediction method to solve the problem of low sunshine calculation efficiency.

[0054] According to an embodiment of the present invention, an embodiment of a sunshine prediction method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0055] In this embodiment, a sunshine prediction method is provided, which can be used in computer devices, such as computers, tablet computers, etc. Figure 1 is a flow chart of a sunshine prediction method according to an embodiment of the present invention. Figure 1As shown, the process includes the following steps:

[0056] Step S11, generating a training data set based on random number parameterization and determining a target sunshine prediction model.

[0057] Random number parameterization generation refers to random sampling multiple times between the preset maximum and minimum values ​​to determine multiple sunlight analysis scenarios. For example, the maximum and minimum values ​​of the latitude of the target area, the maximum and minimum values ​​of the height, width and length of the buildings in the target area, and the maximum and minimum values ​​of the height of the target space point are set. Random sampling is performed K times between these maximum and minimum values ​​to obtain the latitude of K training areas, the height, width and length of K virtual buildings, and the height of K virtual space points, which correspond to K virtual sunlight analysis scenarios. Each scenario corresponds to the building information in the area, the regional geographic information, and the position of the virtual space point, thereby obtaining a training data set.

[0058] The model is trained based on the training data set to obtain a target sunshine prediction model. The specific training method is not limited here.

[0059] Step S12, obtaining geographic information of the target area, geometric attribute information of each building in the target area, and three-dimensional coordinates of the target space point.

[0060] The target area refers to the design area, that is, the area where the architect needs to carry out architectural design. The geographic information of the target area includes the location information of the site where the building is located. The location information of the target area can be latitude. There is no restriction on the height and number of target space points in the target area. There can be multiple target space points at the same height. The buildings in the target area are used to study the shielding of the target space points. The geometric attribute information of the building may include the building size, building location, etc. The building size may include the height, length, and width of the building. The three-dimensional coordinates of the target space point can be determined according to the position of the target space point, and the specific needs need to be determined according to the established three-dimensional coordinate system. The coordinates of the building can be obtained according to the building position and the established coordinate system. For a two-dimensional coordinate system with the bottom surface as a plane, the position of the building can refer to the coordinates of the center point of the ground of the building.

[0061] The geometric attribute information of each building in the target area and the position of the target space point can be considered as the attribute information of the target area, which can be obtained from the design data when performing architectural design. The three-dimensional coordinates of the target space point in the coordinate system are calculated based on the established coordinate system and the position of the target space point.

[0062] by Figure 2For example, there are 311 buildings B and 312 calculation points in the target area. The calculation point 312 is the target space point. From the attribute information of the building, it can be determined that the length of building B (311) is L meters, the width is W meters, and the height is H meters. The center point of the bottom surface of building B is taken as the origin, the due east direction is the x direction, the due north direction is the y direction, and the vertical upward direction is the z direction. The unit length is 1 meter. In the local 3D rectangular coordinate system, it is known that the calculation point 312 is x0, y0, and z0 meters away from the center point of the bottom surface of building B in the due east, due north, and vertical upward directions, respectively. Then the coordinates of the calculation point 312 in the local 3D rectangular coordinate system are (x0, y0, z0). The height of the target space point is determined by the three-dimensional coordinates of the target space point, and the height is z0.

[0063] Step S13, performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine the first geometric parameters corresponding to the building and the second coordinate parameters corresponding to the target space point.

[0064] Due to geometric similarity, if the buildings are enlarged in proportion, their shadows under sunlight will also be enlarged in the same proportion. Therefore, the time when the point (x, y) is blocked by a building with a height of H, a length of L and a width of W is the same as the time when the point (2x, 2y) is blocked by a building with a height of 2H, a length of 2L and a width of 2W. For this reason, in this embodiment, the building geometric attribute information and the three-dimensional coordinates of the target space point in the three-dimensional coordinate system can be converted into a plane coordinate system for processing. The relative size of the building and the relative position relationship between the building and the target space point can be obtained through the geometric coordinate transformation. The specific calculation method is not limited in this embodiment.

[0065] In addition, the above-mentioned geometric coordinate transformation has the effect of data normalization. Data normalization is a method of feature scaling and a key step in data preprocessing. Based on the normalization requirements of deep learning, the different meanings of the input parameters of the neural network will lead to different orders of magnitude. Therefore, it is necessary to normalize the attribute information of the building and the three-dimensional coordinates of the target space point. After normalization, the geometric attribute information of the building is converted into the first geometric parameter, and the three-dimensional coordinates of the target space point are converted into the second coordinate parameter.

[0066] Step S14: for each building, the first geometric parameter, the second coordinate parameter and the geographic information of the target area are respectively input into the target sunshine prediction model to determine the total duration and the start time of the shielding of the target space point by each building.

[0067] The target sunshine prediction model is an artificial neural network model, which is used to predict the starting time and total time of the target space point being shaded by a building from the start time to the end time of the analysis. The input of the model includes the first geometric parameter corresponding to the building, the second coordinate parameter of the target space point, and the geographic information of the target area. The first geometric parameter reflects the size and location information of the building, and the second coordinate parameter reflects the location information of the target space point. Before using the target sunshine prediction model for sunshine prediction, the model needs to be trained.

[0068] The target sunshine prediction model includes an input layer, a hidden layer and an output layer, wherein the hidden layer can have multiple layers, and the specific number of layers can be set according to actual needs, and the output layer includes 2 nodes. When the target sunshine prediction model is used for prediction, the first geometric parameter, the second coordinate parameter and the geographic information of the target area are used as the input data of the input layer. After the calculation of the hidden layer, the two nodes of the output layer output the total duration and the start time of the shielding of the target spatial point by each building. The output result specifically refers to the total duration and the start time of the shielding of a certain spatial point by a certain building in the target area.

[0069] Step S15, based on the total duration that the target space point is shielded by each building and the shielding start time, the total duration that the target space point is shielded is obtained.

[0070] For the M buildings contained in the target area, repeat the above steps S13 to S14 M times to obtain the total duration and the start time of the shielding of the target space point by each building in the target area. If the target area contains N target space points, for each target space point, calculate the total duration and the start time of shielding by any building respectively, and obtain the total duration of shielding of each target space point.

[0071] In this step, if there is only one target space point, the total duration of the target space point being obscured refers to the total duration of the point being obscured by all buildings; if there is more than one target space point, the total duration of the target space point being obscured refers to the total duration of any target space point being obscured.

[0072] Step S16, determining the sunshine duration of the target space point based on the total sunshine duration of the target area and the total shielding duration of the target space point.

[0073] The sunshine duration of a target space point refers to the total time that any target space point receives sunshine. The sunshine duration of a target space point is equal to the total sunshine duration of the target area minus the total time that the target space point is shaded.

[0074] The total sunshine duration of the target area is equal to the difference between the analysis end time and the analysis start time. The analysis start time and the analysis end time can be set to the sunrise time and the sunset time respectively, or can be set according to the climate.

[0075] The sunshine prediction method provided in this embodiment first generates a training data set based on random number parameterization to determine the target sunshine prediction model; obtains the geographic information of the target area, the geometric attribute information of each building in the target area, and the three-dimensional coordinates of the target space point, performs geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point, obtains the first geometric parameter corresponding to the building, and the second coordinate parameter corresponding to the target space point, inputs the first geometric parameter, the second coordinate parameter and the geographic information of the target area into the target sunshine prediction model, and obtains the total duration of the target space point being shaded by each building and the start time of the shade. Then, the total duration of the target space point being shaded is calculated, and the sunshine duration of the target space point is determined based on the total sunshine duration of the target area and the total sunshine duration of the target space point being shaded. The method performs sunshine prediction through neural network fitting, effectively improving the calculation efficiency; and uses geometric coordinate transformation to eliminate the influence of invalid training data, while also having the effect of normalizing the model input data, reducing the difficulty of neural network convergence and fitting error; uses random number parameterization to generate training data sets, does not require a fixed data set, and saves storage space.

[0076] In this embodiment, a sunshine prediction method is provided, which can be used in computer devices, such as computers, tablet computers, etc. The geometric attribute information of a building includes the building size, and the building size includes the length, width and height. Figure 3 is a flow chart of a sunshine prediction method according to an embodiment of the present invention. Figure 3 As shown, the process includes the following steps:

[0077] Step S21, generating a training data set based on random number parameterization and determining a target sunshine prediction model.

[0078] For details, please see Figure 1 Step S11 of the illustrated embodiment will not be described in detail here.

[0079] Step S22, obtaining geographic information of the target area, geometric attribute information of each building in the target area, and three-dimensional coordinates of the target space point.

[0080] For details, please see Figure 1 Step S12 of the illustrated embodiment will not be described in detail here.

[0081] Step S23, performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine the first geometric parameters corresponding to the building and the second coordinate parameters corresponding to the target space point.

[0082] Specifically, step S23 includes the following steps:

[0083] Step S231, subtract the height of the target space point from the sum of the length, width and height of the building to determine the reference size.

[0084] by Figure 2 For example, there are 311 buildings B and 312 calculation points in the target area. The calculation point 312 is the target space point. From the attribute information of the building, it can be determined that the length of building B (311) is L meters, the width is W meters, and the height is H meters. The center point of the bottom surface of building B is taken as the origin, the due east direction is the x direction, the due north direction is the y direction, and the vertical upward direction is the z direction. The unit length is 1 meter. In the local 3D rectangular coordinate system, it is known that the calculation point 312 is x0, y0, and z0 meters away from the center point of the bottom surface of building B in the due east, due north, and vertical upward directions, respectively. Then the coordinates of the calculation point 312 in the local 3D rectangular coordinate system are (x0, y0, z0). The height of the target space point is determined by the three-dimensional coordinates of the target space point, and the height is z0.

[0085] First, the reference size is calculated. The reference size is a standard scale in this embodiment. The reference size Norm_L meters is defined as the length 313 of building B plus the width 314 of building B plus the height 315 of building B minus the height 318 of the calculation point. That is, the reference size Norm_L = L + W + H - z0.

[0086] Step S232, determining the relative length, relative width and relative height of the building based on the reference size, the three-dimensional coordinates of the target space point and the building size.

[0087] See also Figure 2 and Figure 4 , a local 2D rectangular coordinate system and a local 2D polar coordinate system are established at a 320-degree top-down perspective. The unit lengths of the local 2D rectangular coordinate system and the local 2D polar coordinate system are both reference dimensions, and the unit length origins of the local 2D rectangular coordinate system and the local 2D polar coordinate system are both 0.2 unit lengths due south of the center point 321 of the bottom surface of building B. The x direction of the local 2D rectangular coordinate system is due east, the y direction is due south, the 0° direction of the local 2D polar coordinate system is due east, and the 90° direction is due south.

[0088] The relative length 323 of building B is equal to the length 313 of building B divided by the reference size, i.e., relative length = L / Norm_L;

[0089] The relative width 324 of building B is equal to the width 314 of building B divided by the reference size, that is, relative width = W / Norm_L;

[0090] The relative height of building B is equal to the difference between the height 315 of building B and the height 318 of the calculation point divided by the reference size, that is, (H-z0) / Norm_L, where the calculation point is the target space point.

[0091] Step S233, determining the orientation of the building based on the angle between the building size and the preset orientation.

[0092] In this embodiment, the preset direction is the east direction, and the building orientation is the angle between the length direction 325 of the building B and the east direction.

[0093] In actual use, the orientation of the building can also be defined according to actual needs.

[0094] The first geometric parameters of the building include relative length, relative width, relative height and orientation.

[0095] Step S234, determining a second coordinate parameter of the target space point based on the ratio of the three-dimensional coordinate of the target space point to the reference size, the second coordinate parameter including a two-dimensional polar coordinate.

[0096] Based on the established two-dimensional coordinate system, including the local 2D rectangular coordinate system and the local 2D polar coordinate system, the three-dimensional coordinates of the target space point are mapped into the two-dimensional coordinate system to obtain the two-dimensional coordinates. The two-dimensional coordinates can be rectangular coordinates or polar coordinates. The second coordinate parameters corresponding to the target space point include two-dimensional polar coordinates. In addition, the second coordinate parameters can also be rectangular coordinates. Specifically, when predicting the sunshine of the target space point, the starting time and total time when the target space point is shaded by a single building are more directly related to the polar coordinates of the target space point. On the other hand, plane coordinates can be used to generate space points in the training data set. In actual projects, the target space points are usually collected based on an equidistant grid under the plane coordinate system, so the target space points are usually evenly distributed in the plane coordinate system, rather than in the polar coordinate system. In actual applications, the coordinate form used by the second coordinate parameter can be determined according to actual needs. The calculation method of rectangular coordinates and two-dimensional polar coordinates is as follows:

[0097] Step a1, determining the rectangular coordinates of the target space point based on the ratio of the three-dimensional coordinates of the target space point to the reference size.

[0098] The rectangular coordinates of the target space point (,ry), where rx is equal to the x-coordinate of the target space point in the local 3D rectangular coordinate system divided by the reference size; ry is equal to 0.2 plus the quotient of the y-coordinate of the target space point in the local 3D rectangular coordinate system and the reference size.

[0099] That is, rx=x0 / Norm_, rx=y0 / Norm_+0.2.

[0100] Step a2, based on the rectangular coordinates, calculate the two-dimensional polar coordinates of the target space point.

[0101] The calculation parameters corresponding to the target space point include two-dimensional polar coordinates (,r),

[0102] in,

[0103] It should be noted that in this embodiment, there is no limitation on the method of establishing the coordinate system, and there is no limitation on the method of determining the origin of the coordinate system, which can be set according to actual needs.

[0104] Step S24: for each building, the first geometric parameter, the second coordinate parameter and the geographic information of the target area are respectively input into the target sunshine prediction model to determine the total duration of the target space point being shaded by the building and the start time of the shade.

[0105] For details, please see Figure 1 Step S14 of the illustrated embodiment will not be described in detail here.

[0106] Step S25, based on the total duration that the target space point is shielded by each building and the shielding start time, the total duration that the target space point is shielded is obtained.

[0107] For details, please see Figure 1 Step S15 of the illustrated embodiment will not be described in detail here.

[0108] Step S26, determining the sunshine duration of the target space point based on the total sunshine duration of the target area and the total shielding duration of the target space point.

[0109] For details, please see Figure 1 Step S16 of the illustrated embodiment will not be described in detail here.

[0110] The sunshine prediction method provided by the embodiment of the present invention takes into account the normalization requirements of geometric similarity and deep learning, and performs geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point. The parameters in the three-dimensional space are converted into a two-dimensional plane, and the relative size between the building and the target space point is expressed. The method processes the parameters of the input model to reduce the convergence difficulty of the neural network, reduce the fitting error, and thus improve the calculation efficiency.

[0111] In this embodiment, a sunshine prediction method is provided, which can be used in computer devices, such as computers, tablet computers, etc. Figure 5 is a flow chart of a sunshine prediction method according to an embodiment of the present invention. Figure 5 As shown, the process includes the following steps:

[0112] Step S31, generating a training data set based on random number parameterization and determining a target sunshine prediction model.

[0113] For details, please see Figure 1 Step S11 of the illustrated embodiment will not be described in detail here.

[0114] Step S32, obtaining geographic information of the target area, geometric attribute information of each building in the target area, and three-dimensional coordinates of the target space point.

[0115] For details, please see Figure 1 Step S12 of the illustrated embodiment will not be described in detail here.

[0116] Step S33, performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine the first geometric parameters corresponding to the building and the second coordinate parameters corresponding to the target space point.

[0117] For details, please see Figure 1 Step S13 of the illustrated embodiment will not be described in detail here.

[0118] Step S34, for each building, the first geometric parameter, the second coordinate parameter and the geographic information of the target area are respectively input into the target sunshine prediction model to determine the total duration and the start time of the shielding of the target space point by each building.

[0119] For details, please see Figure 1 Step S14 of the illustrated embodiment will not be described in detail here.

[0120] Step S35, based on the total duration that the target space point is shielded by each building and the shielding start time, the total duration that the target space point is shielded is obtained.

[0121] Specifically, step S35 includes the following steps:

[0122] Step S351, based on the total duration of the target space point being shielded by each building and the shielding start time, calculate the shielding end time of the target space point.

[0123] Here, the total duration of the target space point being shielded by each building and the target space point in the shielding start time refer to any target space point in the target area.

[0124] Let the total time TT that the nth target space point is shaded by the shadow of the mth building be m and the starting time ST m , the corresponding occlusion end time is ET m .

[0125] For M buildings in the target area, the initial total duration sequence of the nth target space point being shaded by the shadow of any building is [TT1, ..., TT m ,……,TTM ], the initial masking start time sequence is [ST1,……,ST m ,……,ST M ], the initial occlusion end time sequence is [ET1,……,ET m ,……,ET M ], where ET m =ST m +TT m .

[0126] Step S352, based on the shielding start time and the shielding end time, obtain the total shielding time of the target space point.

[0127] Here, the occlusion start time includes any occlusion start time in the occlusion start time sequence, and the occlusion end time includes any occlusion end time in the occlusion end time sequence, wherein the minimum occlusion start time and the maximum occlusion end time can be understood as the total duration from the start of occlusion to the end of occlusion, which may include a time period when no building causes occlusion to the target space point. By subtracting this time period, the total duration of the target space point being occluded can be obtained.

[0128] In some implementations, step S352 includes the following steps:

[0129] Step b1, sorting the shading start time and the shading end time respectively, and determining the shading start time sequence and the shading end time sequence of the spatial points.

[0130] For M buildings in the target area, the initial total duration sequence of the nth target space point being shaded by the shadow of any building is [TT1, ..., TT m ,……,TT M ], the initial masking start time sequence is [ST1,……,ST m ,……,ST M ], the initial occlusion end time sequence is [ET1,……,ET m ,……,ET M ].

[0131] The initial sequences are sorted in chronological order, and the sorted masking start time sequence is [ST1', ..., ST m ',……,ST M '], the sorted masking end time sequence is [ET1', ..., ET m ',……,ET M '].

[0132] Step b2, calculating the difference between the maximum moment in the masking end moment sequence and the minimum moment in the masking start moment sequence, and determining the first total duration.

[0133] First total duration = ET M '-ST1'.

[0134] Step b3, calculate the unobstructed time period between the maximum moment in the occlusion end time sequence and the minimum moment in the occlusion start time sequence, and calculate the difference between the first total duration and the unobstructed time period to obtain the total duration of the target space point being obscured.

[0135] For the nth target space point, the total duration of being blocked by M buildings is calculated according to the following formula:

[0136]

[0137] Among them, E.T. M ' indicates the end time when the target space point is covered for the last time, ST1' indicates the start time when the target space point is covered for the first time, ST m+1 ' represents the starting time when the target space point is obscured for the m+1th time, ET m 'Indicates the end time of the target space point being covered for the mth time.

[0138] Step S36, determining the sunshine duration of the target space point based on the total sunshine duration of the target area and the total duration of the target space point being shaded.

[0139] For details, please see Figure 1 Step S16 of the illustrated embodiment will not be described in detail here.

[0140] In some embodiments, Figure 1 Step S11 in the embodiment includes the following steps:

[0141] Step S41, generating features of the training data set based on random number parameterization, the features including a first geometric parameter, a second coordinate parameter and geographic information of the training area.

[0142] Specifically, step S41 includes the following steps:

[0143] Step S411, generating geographic information of the training area, geometric information of virtual buildings in the training area and heights of virtual space points based on random number parameterization; Step S412, determining the first geometric parameters of the virtual building based on the geometric information of the virtual building in the training area and heights of virtual space points.

[0144] Random number parameterization generation refers to random sampling multiple times between the preset maximum and minimum values ​​to determine multiple sunshine analysis scenarios. For example, the maximum and minimum values ​​of the latitude of the target area, the maximum and minimum values ​​of the height, width and length of the buildings in the target area, and the maximum and minimum values ​​of the height of the target space point are set. Random sampling is performed K times between these maximum and minimum values ​​to obtain the latitude of K training areas, the height, width and length of K virtual buildings, and the height of K virtual space points, which correspond to K virtual sunshine analysis scenarios.

[0145] For any sunlight analysis scenario, coordinate transformation is performed to establish a three-dimensional coordinate system and a two-dimensional coordinate system. The two-dimensional coordinate system includes a local 2D rectangular coordinate system and a local 2D polar coordinate system. The relative length, relative width, relative height and orientation of the building are obtained based on the geometric attribute information of the building. The calculation method can be found in the above steps S231, S232 and S233, which will not be repeated here.

[0146] Step S413: determining the second coordinate parameter of the virtual space point in the preset sampling area based on the preset first step length sampling network and the preset second step length sampling network.

[0147] Among them, the first step length is greater than the second step length.

[0148] The first step sampling network and the second step sampling network perform spatial point sampling in the sampling area to determine the second coordinate parameters of each spatial point, that is, the rectangular coordinates or polar coordinates in the two-dimensional coordinate system. The sampling area is a legal sampling area. The specific calculation method of the second coordinate parameter is referred to the above step S234, which will not be repeated here.

[0149] For any sunshine analysis scene, all virtual space points are generated in the established local 2D rectangular coordinate system and local 2D polar coordinate system using the mixed step floating starting point calculation point sampling method. There are multiple virtual space points at one target height.

[0150] Please refer to Figure 6 , the mixed step floating starting point calculation point sampling method is as follows:

[0151] The legal sampling area 550 is an arched area, which can basically cover all coordinate points that will be obscured by building shadows from the start time to the end time of the analysis in any sunlight analysis scenario.

[0152] For example, for the sunshine analysis scenario where the site latitude is between 25°N and 35°N, the analysis starts at 8:00 AM and ends at 16:00 AM, the recommended 550 legal sampling areas are θ∈[25°,35°], r∈[0.2,3.2].

[0153] One or more step sizes are used to sample in a preset site. Here, two step sizes are used as an example, namely a small step size sampling grid 530 and a large step size sampling grid 540. The small step size sampling grid 530 is used for fine sampling, which can basically cover all coordinate points that are covered by building shadows for a total time of more than 2 hours from the start time to the end time of the analysis in any sunlight analysis scenario. The large step size sampling grid 540 is used for coarse sampling in a larger range and needs to be slightly larger than the legal sampling area 550.

[0154] For example, for the sunshine analysis scenario with the site latitude between 25°N and 35°N, the analysis start time is 8:00 AM true solar time on the Great Cold Day, and the analysis end time is 16:00 PM true solar time on the Great Cold Day, the recommended small step sampling grid 530 is x∈[-1,1], y∈[0.2,1.2], and the sampling step is 0.05; the recommended large step sampling grid 540 is x∈[-3,3], y∈[0,3.2], and the sampling step is 0.1.

[0155] The small step sampling grid 530 and the large step sampling grid 540 are randomly offset to obtain an offset small step sampling grid and an offset large step sampling grid. The offset amount of the random offset in the x and y directions is randomly selected within the range of -0.5 to 0.5 sampling steps.

[0156] The sampled virtual space points are all coordinate points in the 550 legal sampling area among the grid points of the offset small-step sampling grid and the offset large-step sampling grid.

[0157] The mixed step-size floating starting point calculation point sampling method is helpful for training artificial neural networks and can improve the prediction accuracy and generalization ability of the trained artificial neural networks.

[0158] The geographic information of the training area, the first geometric parameters of the virtual building and the second coordinate parameters of each virtual space point are combined into the features of the training data set.

[0159] The features of any piece of training data are: [latitude of the site, orientation of building B, relative length of building B, relative width of building B, relative height of building B, spatial point coordinate θ, spatial point coordinate r].

[0160] Step S42, input the features of the training data set into the reference sunshine prediction model, and output the true value of the training data set label.

[0161] The reference sunshine prediction model refers to the model used by sunshine prediction software in the prior art, and is not specifically limited. The real value is calculated by the reference sunshine training model and used for the training of the subsequent target sunshine prediction model. The training data set label includes the total time that the spatial point is shaded by the building and the start time of the shade.

[0162] Step S43, input the features of the training data set into the target sunshine prediction model, and output the predicted value of the training data set label. Figure 7 As shown, the model includes an input layer 200 , a first hidden layer 210 , a second hidden layer 220 , a third hidden layer 230 and an output layer 240 .

[0163] The input layer 200 includes 7 nodes, which respectively represent the latitude 201 of the site, the orientation 202 of building B, the relative length 203 of building B, the relative width 204 of building B, the relative height 205 of building B, the θ coordinate 206 of the spatial point, and the r coordinate 207 of the spatial point. The calculation point in the figure is the spatial point.

[0164] The first hidden layer 210 and the second hidden layer 220 include 16 nodes, and the third hidden layer 230 includes 8 nodes.

[0165] The output layer 240 includes two nodes, which respectively represent the total duration TT of the space point (θ, r) being shaded by the shadow of the building B, and the starting time ST of the space point (θ, r) being shaded by the shadow of the building B.

[0166] Step S44, calculating the loss according to the real value and the predicted value of the label, and updating the parameters of the target sunshine prediction model based on the loss to determine the target sunshine prediction model.

[0167] The traditional sunshine analysis method is used to calculate the total time TT of any spatial point being shaded by the shadow of the building in any sunshine analysis scene. GT and the starting time ST GT , which is the true value output by step S42.

[0168] The loss of the defined model = total time loss + start time loss, where:

[0169] Total time loss = (TT-TT GT ) 2

[0170]

[0171] Wherein, TT represents the predicted total shielding time of (θ, r) output by the target sunshine prediction model being shielded by the shadow of building B, ST represents the starting time of (θ, r) output by the target sunshine prediction model being shielded by the shadow of building B; TT GT is the total time that (θ,r) is shaded by the shadow of building B predicted by the traditional sunshine analysis method, ST GT It is the starting time when (θ, r) is shaded by the shadow of building B as predicted by the traditional sunshine analysis method.

[0172] If the model loss is less than the preset threshold, the model training is completed and the target sunshine prediction model is determined. If the model loss is not less than the preset threshold, the model parameters are adjusted based on the model loss and error back propagation algorithm, and the training data generated in the above steps are deleted from the memory, and the training data and model loss are regenerated and calculated until the loss is less than the preset threshold.

[0173] In the implementation manner of the present invention, a random number parameterization method is adopted to generate training data sets in batches, and training is performed while the data sets are generated. Training and data generation are performed simultaneously. Compared with sampling a large number of fixed training data sets, this method saves more storage space.

[0174] The sunshine prediction method provided by the embodiment of the present invention is based on an artificial neural network and is a "one-step" sunshine prediction method. It significantly reduces the calculation time while ensuring that the prediction accuracy meets the requirements. When the method is used as a sunshine simulation tool, the calculation time is only 1 / 77 of the traditional sunshine prediction method, and the calculation accuracy is 98% of the traditional sunshine prediction method; when the method is used as an evaluation function of a building design optimization tool, the calculation time is only 1 / 54 of the traditional sunshine prediction method, and the calculation accuracy is 96.5% of the traditional sunshine prediction method.

[0175] In this embodiment, a building sunshine prediction device is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0176] This embodiment provides a sunshine prediction device, such as Figure 8 As shown, including:

[0177] A model determination module 81 is used to generate a training data set based on random number parameterization and determine a target sunshine prediction model;

[0178] The data acquisition module 82 is used to acquire the geographic information of the target area, the geometric attribute information of each building in the target area, and the three-dimensional coordinates of the target space point;

[0179] The data processing module 83 is used to perform geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine the first geometric parameter corresponding to the building and the second coordinate parameter corresponding to the target space point;

[0180] A parameter calculation module 84 is used to input the first geometric parameter, the second coordinate parameter and the information of the target area into a target sunshine prediction model for each of the buildings, and determine the total duration and the start time of the shielding of the target space point by the building;

[0181] A first calculation module 85, for obtaining a total time duration of shielding of the target space point based on the total time duration of shielding of the target space point by each building and the shielding start time;

[0182] The second calculation module 86 is used to determine the sunshine duration of the target space point based on the total sunshine duration of the target area and the total duration of the target space point being shaded.

[0183] In some optional implementations, the first calculation module 85 includes:

[0184] An end calculation unit, used for calculating the shading end time of the target space point based on the total duration of the target space point being shaded by each building and the shading start time;

[0185] The total duration calculation unit is used to obtain the total duration of the target space point being covered based on the covering start time and the covering end time.

[0186] In some optional implementations, the total duration calculation unit includes:

[0187] A time sorting subunit, used to sort the shielding start time and the shielding end time respectively, and determine the shielding start time sequence and the shielding end time sequence of the spatial point;

[0188] A first calculation subunit is used to calculate the difference between the maximum moment in the masking end moment sequence and the minimum moment in the masking start moment sequence to determine a first total duration;

[0189] The second calculation subunit is used to calculate the unobstructed time period between the maximum moment in the occlusion end time sequence and the minimum moment in the occlusion start time sequence, and calculate the difference between the first total duration and the unobstructed time period to obtain the total duration of the target space point being obscured.

[0190] In some optional implementations, the geometric attribute information of the building includes building dimensions, and the building dimensions include length, width, and height. The data processing module 83 includes:

[0191] A reference size calculation unit, used for determining a reference size by subtracting the height of the target space point from the sum of the length, width and height of the building, wherein the height of the target space point is determined by the three-dimensional coordinates of the target space point;

[0192] a relative size determination unit, configured to determine a relative length, a relative width and a relative height of the building based on the reference size, the three-dimensional coordinates of the target space point and the building size;

[0193] An orientation determining unit, configured to determine the orientation of a building based on an angle between the building size and a preset orientation, wherein the first geometric parameter of the building includes a relative length, a relative width, a relative height and an orientation;

[0194] The second parameter calculation unit is used to determine the second coordinate parameter of the target space point based on the ratio of the three-dimensional coordinate of the target space point to the reference size, wherein the second coordinate parameter includes a two-dimensional polar coordinate.

[0195] In some optional implementations, the model determination module 81 includes:

[0196] A data generation unit, configured to generate features of a training data set based on random number parameterization, wherein the features include a first geometric parameter, a second coordinate parameter, and geographic information of a training area;

[0197] A first output unit, used to input the features of the training data set into a reference sunshine prediction model and output a true value of a label of the training data set;

[0198] A second output unit is used to input the features of the training data set into a target sunshine prediction model and output a predicted value of a label of the training data set;

[0199] The model determination unit is used to calculate the loss according to the real value and the predicted value of the label, and update the parameters of the target sunshine prediction model based on the loss to determine the target sunshine prediction model.

[0200] In some optional implementations, the data generating unit includes:

[0201] A data generation subunit, for generating geographic information of a training area, geometric information of virtual buildings in the training area, and heights of virtual space points based on random number parameterization;

[0202] A first generating subunit, configured to determine a first geometric parameter of the virtual building based on geometric information of the virtual building in the training area and a height of a virtual space point;

[0203] The second sampling subunit is used to determine the second coordinate parameter of the virtual space point in a preset sampling area based on a preset first step length sampling network and a second step length sampling network, wherein the first step length is greater than the second step length.

[0204] The sunshine prediction device in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0205] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0206] The embodiment of the present invention also provides a computer device having the above Figure 8 The sunshine prediction device shown.

[0207] See also Fig. 9 , Fig. 9 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Fig. 9 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Fig. 9 A processor 10 is taken as an example.

[0208] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0209] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0210] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created by the use of a computer device based on the presentation of a small program landing page, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0211] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0212] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0213] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0214] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A sunshine prediction method, characterized in that: The method comprises: Generate training data set based on random number parameterization and determine target sunshine prediction model; Obtaining geographic information of the target area, geometric attribute information of each building in the target area, and three-dimensional coordinates of the target space point; Performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point; For each of the buildings, the first geometric parameter, the second coordinate parameter and the geographic information of the target area are respectively input into a target sunshine prediction model to determine the total duration and the start time of the shielding of the target spatial point by each of the buildings; Based on the total duration of shielding of the target space point by each building and the shielding start time, the total duration of shielding of the target space point is obtained; The sunshine duration of the target space point is determined based on the total sunshine duration of the target area and the total duration of shielding of the target space point.

2. The method according to claim 1, characterized in that The obtaining of the total duration of shielding of the target space point based on the total duration of shielding of the target space point by each building and the shielding start time includes: Calculate the shading end time of the target space point based on the total duration of the target space point being shaded by each building and the shading start time; Based on the shielding start time and the shielding end time, the total duration of shielding of the target space point is obtained.

3. The method according to claim 2, characterized in that The obtaining, based on the shielding start time and the shielding end time, of the total shielding duration of the target space point includes: The shielding start time and the shielding end time are sorted respectively to determine the shielding start time sequence and the shielding end time sequence of the spatial point; Calculate the difference between the maximum time in the masking end time sequence and the minimum time in the masking start time sequence to determine a first total duration; The unobstructed time period between the maximum moment in the obstruction end time sequence and the minimum moment in the obstruction start time sequence is calculated, and the difference between the first total duration and the unobstructed time period is calculated to obtain the total duration of the target space point being obscured.

4. The method according to claim 1, characterized in that: The geometric attribute information of the building includes building dimensions, and the building dimensions include length, width, and height. The geometric coordinate transformation is performed on the geometric attribute information of the building and the three-dimensional coordinates of the target space point to determine a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point, including: Determine the reference size by subtracting the height of the target space point from the sum of the length, width and height of the building, where the height of the target space point is determined by the three-dimensional coordinates of the target space point; Determine the relative length, relative width and relative height of the building based on the reference size, the three-dimensional coordinates of the target space point and the building size; Determining the orientation of the building based on the angle between the building size and the preset orientation, wherein the first geometric parameter of the building includes a relative length, a relative width, a relative height and an orientation; A second coordinate parameter of the target space point is determined based on a ratio of the three-dimensional coordinate of the target space point to the reference size, wherein the second coordinate parameter includes a two-dimensional polar coordinate.

5. The method according to claim 1, characterized in that The generating of a training data set based on random number parameterization and determining a target sunshine prediction model comprises: Generate features of the training data set based on random number parameterization, the features including a first geometric parameter, a second coordinate parameter, and geographic information of a training area; Input the features of the training data set into a reference sunshine prediction model, and output the true value of the training data set label; Input the features of the training data set into a target sunshine prediction model, and output a predicted value of the training data set label; The loss is calculated according to the true value and the predicted value of the label, and the parameters of the target sunshine prediction model are updated based on the loss to determine the target sunshine prediction model.

6. The method according to claim 5, characterized in that The features of the training data set generated based on random number parameterization include: Generate geographic information of the training area, geometric information of virtual buildings in the training area, and heights of virtual space points based on random number parameterization; Determine a first geometric parameter of the virtual building based on the geometric information of the virtual building in the training area and the height of the virtual space point; Based on a preset first step length sampling network and a second step length sampling network, a second coordinate parameter of the virtual space point is determined in a preset sampling area, and the first step length is greater than the second step length.

7. A sunshine prediction device, characterized in that: The device comprises: A model determination module is used to generate a training data set based on random number parameterization and determine a target sunshine prediction model; A data acquisition module is used to acquire the geographic information of the target area, the geometric attribute information of each building in the target area, and the three-dimensional coordinates of the target space point; A data processing module, used for performing geometric coordinate transformation on the geometric attribute information of the building and the three-dimensional coordinates of the target space point, and determining a first geometric parameter corresponding to the building and a second coordinate parameter corresponding to the target space point; A parameter calculation module, for inputting the first geometric parameter, the second coordinate parameter and the geographic information of the target area into a target sunshine prediction model for each of the buildings, and determining a total duration and a start time of shielding of the target spatial point by each of the buildings; A first calculation module is used to obtain the total time duration of the target space point being shielded based on the total time duration of the target space point being shielded by each building and the shielding start time; The second calculation module is used to determine the sunshine duration of the target space point based on the total sunshine duration of the target area and the total duration of the target space point being shaded.

8. The device according to claim 7, characterized in that The model determination module comprises: A feature generation unit, configured to generate features of a training data set based on random number parameterization, wherein the features include a first geometric parameter, a second coordinate parameter, and geographic information of a training area; A true value acquisition unit, used to input the features of the training data set into a reference sunshine prediction model and output a true value of the training data set label; A prediction value acquisition unit, used to input the features of the training data set into a target sunshine prediction model and output a prediction value of the training data set label; The target model determination unit is used to calculate the loss according to the real value and the predicted value of the label, and update the parameters of the target sunshine prediction model based on the loss to determine the target sunshine prediction model.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the sunshine prediction method according to any one of claims 1 to 6 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the sunshine prediction method according to any one of claims 1 to 6.

Images