CN113570156A - Thermal environment prediction model and thermal environment prediction method based on agricultural facilities - Google Patents

Abstract

The invention provides a thermal environment prediction model and a thermal environment prediction method based on agricultural facilities. The model includes: a first memory module, which is used to perform feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment. The sequence information includes: environmental parameters and target parameters of the thermal environment; the second memory module is used to extract the short-term spatiotemporal characteristics of the thermal environment based on the long-term spatiotemporal characteristics; the autoregressive module is used to determine the initial prediction result based on the historical information of the target parameters ; The output module is used to output the final prediction result of thermal environment based on long-term spatiotemporal features, short-term spatiotemporal features and initial prediction results. The invention is used to solve the defect that the prior art does not combine the characteristics of the thermal environment of agricultural facilities, resulting in poor effect in the long-term prediction task of the thermal environment.

Description

Thermal environment prediction model and thermal environment prediction method based on agricultural facilities

technical field

The invention relates to the technical field of agricultural facilities, in particular to a thermal environment prediction model and a thermal environment prediction method based on agricultural facilities.

Background technique

The prediction methods for thermal environment in facility agriculture (greenhouse, factory aquaponics, soilless culture, recirculating aquaculture) can be divided into mechanism model prediction, machine learning model prediction and time series model prediction. The latter two methods are called "black box models" because they do not pay attention to the laws and principles of physics in the system. The application of these types of methods in mechanism prediction modeling will be introduced in detail below.

Mechanism models use biological and physical principles to quantitatively analyze relevant factors in a system, and establish equilibrium equations through the laws of conservation of energy and mass. The mechanism model includes a static model and a dynamic model: the static model is also called the steady-state model; the dynamic model is usually described by a differential equation, which describes the changing law of the system. The earliest static model of greenhouse environment was established by scholar Businger, which laid the foundation for future research. Although static models are easy to implement, their accuracy is low. Therefore, Japanese scholars Takakura et al. established the first relatively complete kinetic model for the unheated single-layer glass greenhouse, which comprehensively described the heat and moisture transfer process in the greenhouse.

With the continuous improvement of the mechanism model, it also brings convenience to the prediction and regulation of the agricultural facility environment. However, agricultural systems tend to be time-varying due to the large number of parameters and physical variables used in mechanistic models. Therefore, it is difficult to tune in development and practice, and more and more researchers start to focus on data-based models. And, with the development of artificial intelligence, computing performance and data productivity have been further improved, thus promoting the application of machine learning models in agricultural environment modeling.

Although the existing technology can achieve time series prediction through machine learning, it does not combine the characteristics of the thermal environment of agricultural facilities, and cannot achieve satisfactory results in the long-term thermal environment prediction task. Therefore, how to achieve accurate prediction of the thermal environment of agricultural facilities is an important issue to be solved urgently in the industry.

SUMMARY OF THE INVENTION

The invention provides a thermal environment prediction model and a thermal environment prediction method based on agricultural facilities, which are used to solve the defect that the existing technology does not combine the characteristics of the thermal environment of agricultural facilities, resulting in poor effect in the long-term prediction task of the thermal environment, and realizes the improvement of the thermal environment. Precise predictions.

The present invention provides a thermal environment prediction model based on agricultural facilities, comprising: a first memory module for performing feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, where the time series information includes: environmental parameters and target parameters of the thermal environment;

a second memory module, configured to extract short-term spatiotemporal features of the thermal environment based on the long-term spatiotemporal features;

The autoregressive module is used to determine the initial prediction result based on the historical information of the target parameter;

An output module, configured to output the final prediction result of the thermal environment based on the long-term spatiotemporal features, the short-term spatiotemporal features and the initial prediction result.

According to a thermal environment prediction model based on agricultural facilities provided by the present invention, the second memory module includes: an attention unit and at least two second memory units;

The attention unit is used for connecting the first memory module and the second memory module, for adjusting the long-term spatiotemporal features, and transmitting the adjusted long-term spatiotemporal features to the second memory unit.

According to a thermal environment prediction model based on agricultural facilities provided by the present invention, the first memory module includes at least two first memory units;

The attention unit is specifically configured to calculate the attention weight corresponding to each of the first memory units based on the long-term space-time feature and the short-term space-time feature after the long-term space-time feature is obtained, and based on the attention force weight, adjust the long-term spatiotemporal feature, and transmit the adjusted long-term spatiotemporal feature to the second memory unit.

According to a thermal environment prediction model based on agricultural facilities provided by the present invention, the second memory unit is configured to determine the short-term spatiotemporal characteristics based on the adjusted long-term spatiotemporal characteristics and the historical information.

According to a thermal environment prediction model based on agricultural facilities provided by the present invention, the second memory module is a recurrent neural network RNN, and the second memory unit includes: a long short-term memory network LSTM.

According to a thermal environment prediction model based on agricultural facilities provided by the present invention, the first memory unit includes: a time-domain convolutional network TCN.

According to a thermal environment prediction model based on agricultural facilities provided by the present invention, the output module is specifically configured to sum the long-term spatiotemporal features, the short-term spatiotemporal features, and the initial prediction results to obtain the final forecast result.

The present invention also provides a thermal environment prediction method based on a thermal environment prediction model of agricultural facilities, comprising:

Perform feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, where the time series information includes: environmental parameters and target parameters of the thermal environment;

extracting short-term spatiotemporal features of the thermal environment based on the long-term spatiotemporal features;

determining an initial prediction result based on the historical information of the target parameter;

Based on the long-term spatiotemporal features, the short-term spatiotemporal features, and the initial prediction result, a final prediction result of the thermal environment is output.

In the thermal environment prediction method provided by the present invention, the short-term spatiotemporal features of the thermal environment are extracted based on the long-term spatiotemporal features, including:

The long-term spatiotemporal features are adjusted, and the short-term spatiotemporal features are determined based on the adjusted long-term spatiotemporal features and the historical information.

In the thermal environment prediction method provided by the present invention, the output of the final prediction result of the thermal environment based on the long-term spatiotemporal characteristics, the short-term spatiotemporal characteristics and the initial prediction result includes:

The long-term spatiotemporal feature, the short-term spatiotemporal feature, and the initial prediction result are summed to obtain the final prediction result.

In the thermal environment prediction model and thermal environment prediction method based on agricultural facilities provided by the present invention, the thermal environment prediction result is obtained through the thermal environment prediction model, wherein the thermal environment prediction model includes: a first memory module, a second memory module, an automatic regression module and output module. The present invention uses the first memory module to perform feature extraction on the time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, wherein the time series information includes: environmental parameters and target parameters of the thermal environment. It can be seen that the thermal environment of the present invention The environmental prediction model effectively obtains the long-term spatial-temporal characteristics of the thermal environment based on the relevant parameters of the thermal environment; the second memory module, based on the long-term spatial-temporal characteristics of the thermal environment, extracts the short-term spatial-temporal characteristics of the thermal environment. The short-term spatiotemporal characteristics of the environment; the autoregressive module is used to determine the initial prediction results based on the historical information of the target parameters; the output module is used to output the final prediction results of the thermal environment based on the long-term spatiotemporal characteristics, short-term spatiotemporal characteristics and the initial prediction results, It can be seen that the present invention fully considers the long-term space-time characteristics and short-term space-time characteristics of the thermal environment, as well as the spatial relationship between the long-term space-time characteristics and the short-term space-time characteristics, so that the present invention can better adapt to the thermal environment prediction of agricultural facilities, and finally achieve accurate Thermal environment prediction for agricultural facilities.

Description of drawings

In order to explain the present invention or the technical solutions in the prior art more clearly, the following will briefly introduce the accompanying drawings that need to be used in the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are the For some embodiments of the invention, for those of ordinary skill in the art, other drawings can also be obtained according to these drawings without any creative effort.

Fig. 1 is one of the structural representations of the thermal environment prediction model provided by the present invention;

Fig. 2 is the second structural schematic diagram of the thermal environment prediction model provided by the present invention;

Fig. 3 is the third structural schematic diagram of the thermal environment prediction model provided by the present invention;

4 is a schematic flowchart of a thermal environment prediction method provided by the present invention;

FIG. 5 is a schematic structural diagram of an electronic device provided by the present invention.

Detailed ways

In order to make the objectives, technical solutions and advantages of the present invention clearer, the technical solutions in the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention. , not all examples. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

The thermal environment prediction model based on agricultural facilities of the present invention will be described below with reference to FIGS. 1 to 3 .

The embodiment of the present invention provides a thermal environment prediction model based on agricultural facilities. The schematic diagram of the structure of the model is shown in FIG. 1 , and the model specifically includes:

The first memory module 101 is configured to perform feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal features of the thermal environment, where the time series information includes: environmental parameters and target parameters of the thermal environment.

The second memory module 102 is configured to extract short-term spatiotemporal features of the thermal environment based on long-term spatiotemporal features.

The autoregression module 103 is configured to determine the initial prediction result based on the historical information of the target parameter.

The output module 104 is configured to output the final prediction result of the thermal environment based on the long-term spatiotemporal characteristics, the short-term spatiotemporal characteristics and the initial prediction result.

Among them, the long-term spatiotemporal features include: seasonal information, change trend information, and spatiotemporal correlations of the thermal environment; short-term spatiotemporal features include: short-term volatility of the thermal environment.

Specifically, the present invention is mainly aimed at the task of predicting the time series information of the thermal environment of agricultural facilities. The environmental parameters include: outdoor temperature, outdoor humidity, soil temperature, soil moisture, atmospheric pressure, indoor light, indoor carbon dioxide (CO ₂ ) concentration, indoor temperature, indoor humidity, circulating water temperature, and the like. Among them, the environmental parameters include: exogenous variables and target variables. The target variable corresponds to the final prediction result.

Below, take outdoor temperature, outdoor humidity, soil temperature, soil moisture, atmospheric pressure, indoor light, indoor carbon dioxide (CO ₂ ) concentration, indoor temperature, indoor humidity, circulating water temperature, etc. as exogenous variables, and circulating water temperature as the target variable as an example Be explained:

Among them, the exogenous variables and target variables are represented by u _t ∈ R ^N and v _t ∈ R respectively, and the two together constitute the input parameters of the model z _t ={u _t ;v _t }. The goal of the present invention is to predict the target variable of H time steps in the future by using the environmental parameters of the historical T time steps, that is, the time series information. That is, the thermal environment prediction model establishes the mapping function f(·) between the exogenous variable and the target variable, see formula (1)

which is

Y=f(X) (1)

预测的目标变量为Y，Y＝{v_T+1；v_T+2；…；v_T+H}∈R^H，T为常量，H为常量where the input variable is X,

The predicted target variable is Y, Y={v _T+1 ; v _T+2 ;…;v _T+H }∈R ^H , T is a constant, H is a constant

In a specific embodiment, the first memory module 101 includes at least two first memory units, and the first memory units include: a Temporal Convolutional Network (TCN for short).

Specifically, due to the characteristics of strong coupling, large inertia and nonlinearity of the thermal environment of facility agriculture, multivariate data and long-term time series need to be comprehensively considered when creating a model. Therefore, the first memory module of the present invention adopts the TCN model. Among them, the TCN model ensures that future information is not leaked through causal convolution. Furthermore, the TCN model can efficiently obtain spatiotemporal features of arbitrary length through deeper layers and dilated convolutions.

Therefore, the present invention can capture long-term time series information, and extract long-term spatiotemporal features of the thermal environment based on the long-term time series information.

Specifically, as shown in FIG. 2, each TCN model includes two causal convolution TCN sub-modules 201, and each TCN sub-module includes: an expanded convolution layer 2011, a weight normalization layer 2012, an activation function (Rectified Linear Units, referred to as ReLU) 2013 and Dropout layer 2014. Causal convolution means that for a neuron with time step t, it can only use the information before the t-th time step in the previous layer to perform convolution, which can effectively ensure timing information leakage. For the convolution kernel f of size 1×k, after adding the dilated convolution factor d, the receptive field of the feature map Z shows a multiple increase compared with the general convolution module, see formula (2). Further, through weight normalization 2012, ReLU activation function 2013 and Dropout layer 2014, the model has good generalization performance. Finally, residual connections are deployed for the model via 1×1 convolutions 202.

Among them, k is a constant, t is a constant, i is a constant, d is the dilated convolution factor, f is the convolution kernel, and Z is the result of each layer of convolution calculation.

Based on the above-mentioned TCN model, the present invention extracts the long-term space-time features of the thermal environment, the long-term space-time features at time t are represented by m _i , and the long-term space-time features output by the TCN model are M={m ₁ ,m ₂ ,...,m _T }∈ RT ^×p .

In a specific embodiment, the second memory module 102 includes: an attention unit and at least two second memory units; the attention unit is used to connect the first memory module 101 and the second memory module 102 and used to adjust the long-term spatiotemporal features. , and transmit the adjusted long-term spatiotemporal features to the second memory unit.

In a specific embodiment, the second memory module 102 is a recurrent neural network (Recurrent Neural Network, RNN for short), and the second memory unit includes: a long short-term memory network LSTM.

The invention adopts RNN network to effectively capture short-term volatility in thermal environment.

In a specific embodiment, the attention unit is specifically used to calculate the attention weight corresponding to each first memory unit based on the long-term space-time feature and the short-term space-time feature after obtaining the long-term space-time feature, and adjust the long-term space-time based on the attention weight. features, and transmits the adjusted long-term spatiotemporal features to the second memory unit.

作为注意力单元的查询目标(Query)，TCN模型的输出结果m_i∈R^p作为注意力单元的查询键值(Key)，进而，通过

和m_i计算注意力权重，见公式(3)和公式(4)：Specifically, the present invention effectively fuses long-term spatiotemporal features and short-term spatiotemporal features, so that the RNN network can adaptively select a valuable first memory unit for reasoning. For example, at time t, the output result of the hidden layer unit of the LSTM network

As the query target (Query) of the attention unit, the output result of the TCN model m _i ∈ R ^p is used as the query key value (Key) of the attention unit, and then, through

and m _i to calculate the attention weights, see Equation (3) and Equation (4):

W^attn∈R^m×2q、U^attn∈R^m×p和v^attn∈R^m是注意力单元的训练参数，

作为注意力单元的查询目标，m_i作为注意力单元的查询键值。公式(4)用于将计算出的对于每个查询键值的中间能量值

转换为注意力权重

Among them, the hidden units of the LSTM network output d _t-1 and s _t-1 . By splicing d _t-1 and s _t-1 , we get

W ^attn ∈ R ^m×2q , U ^attn ∈ R ^m×p and v ^attn ∈ R ^m are the training parameters of the attention unit,

As the query target of the attention unit, _mi is the query key value of the attention unit. Formula (4) is used to calculate the intermediate energy value for each query key value

Convert to attention weights

Among them, the attention weight can reflect the importance of the i-th first memory unit to the model's prediction target. After the model adaptively selects the first memory unit, the calculation of the LSTM network context information c _t at the t-th time step is shown in formula (5):

Among them, ct is an input parameter of the LSTM network, and _T is a constant.

Specifically, _ct is the adjusted long-term spatiotemporal feature, and ct is transmitted to the _LSTM network.

In a specific embodiment, the second memory module 102 is specifically configured to determine the short-term spatiotemporal characteristics based on the adjusted long-term spatiotemporal characteristics and historical information.

Specifically, the input information of the LSTM network can be obtained by splicing the long-term memory context information c _t-1 processed by the attention unit with the historical information v _t-1 information of the target variable short time window, see formula (6) and formula (7):

和

是训练参数，

表示v_t-1和c_t-1的拼接结果，f_LSTM表示LSTM网络的前向传播过程，并且

用于表示LSTM网络的输出结果，即短期时空特征。in,

and

are the training parameters,

represents the concatenation result of v _t-1 and c _t-1 , f _LSTM represents the forward propagation process of the LSTM network, and

Used to represent the output of the LSTM network, that is, short-term spatiotemporal features.

输入至LSTM网络，输出

Specifically, will

Input to LSTM network, output

In a specific embodiment, the output module 104 is specifically configured to sum the long-term spatiotemporal features, the short-term spatiotemporal features and the initial prediction result to obtain the final prediction result.

Specifically, the acquisition method of the initial prediction result is shown in formula (8):

和b^ar是训练参数，

表示根据历史信息{v_t-T+1,v_t-T+2,…,v_t}的单步初始预测结果。基于单步初始预测结果，得到初始预测结果

in,

and ^bar are training parameters,

Represents the single-step initial prediction result based on historical information {v _t-T+1 ,v _t-T+2 ,…,v _t }. Based on the single-step initial prediction results, the initial prediction results are obtained

The autoregressive module of the present invention is used to solve the generalization problem of the prediction model, and improves the generalization robustness of the prediction model when the prediction model has high accuracy for thermal environment prediction.

Specifically, for the hidden results output by the TCN model and the RNN model, the long-term spatiotemporal features extracted by the TCN model and the short-term spatiotemporal features output by the RNN model are projected to the output latitude required by the thermal environment prediction task through a set of affine transformations, see Equation (9) and Equation (10):

表示LSTM单元在t时刻输出的短期时空特征。where W ^tcn ∈ R ^H×T , U ^tcn ∈ R ^p , b ^tcn ∈ R ^H , W ^rnn ∈ R ^H×2q and b ^rnn ∈ R ^H are training parameters. M={m ₁ ,m ₂ ,…,m _T }∈R ^T×p represents the long-term spatiotemporal features output by the TCN model,

Represents the short-term spatiotemporal features output by the LSTM unit at time t.

Finally, the local outputs of the TCN model, the RNN model and the autoregressive module 103 are added to obtain the multi-step prediction result of the thermal environment, as shown in formula (11). In addition, since the model is smooth and differentiable, the parameters in the prediction model can be trained by the back-propagation algorithm, see equation (12):

和真实热环境数据Y⁽ⁱ⁾之间的L2范数。where Θ represents all the training parameters in the model and N is the batch size in the mini-batch gradient descent algorithm. ||·|| ₂ represents the final prediction result predicted by the model

and the L2 norm between the real thermal environment data Y ⁽ⁱ⁾ .

In addition, after the training of the prediction model is completed, the prediction accuracy of the model needs to be verified, as follows:

In the present invention, different evaluation strategies are proposed from two aspects of global performance index and local performance index. Among them, the prediction effect of multiple future time steps is considered as the global performance, focusing on the overall prediction accuracy of the model and evaluated with root mean square error (RMSE), mean absolute error (MAE) and mean absolute percent error (MAPE) , see Equation (13), Equation (14) and Equation (15):

Second, the fit of each prediction step (i.e., single-step evaluation in multi-step prediction results) is considered as a local performance, effectively reflecting how well the prediction results match in real production. The relative square root error (RRSE) and empirical correlation coefficient (CORR) were used to measure the local performance of the model, see equations (16) and (17):

是模型预测值，y是真实测量值，则以上指标的定义如下：in,

is the model predicted value and y is the actual measured value, the above indicators are defined as follows:

Global performance metrics:

Local performance indicators:

Among them, i is a constant representing any number of data samples, h represents the step size of multi-step forecasting of time series in the forecasting model, N is a constant representing the total number of data samples, and Ω represents the set of data samples and multi-step forecasting results.

Among them, for the RMSE, MAE, MAPE and RRSE indicators, the larger the better; for the CORR indicator, the smaller the better.

Below, the thermal environment prediction model of the present invention is described in detail through FIG. 3:

First, the exogenous variable u and the target variable v are connected and serialized through the long historical information time window T as the input of the multi-layer TCN model, and the long-term _{spatiotemporal} features mi are obtained.

输入到注意力单元，得到注意力单元输出c_t，其中，

为LSTM网络的隐藏层输出，即，短期时空特征。Then, [m ₁ ,m ₂ ,...,m _T ] and

Input to the attention unit, and get the attention unit output c _t , where,

output for the hidden layer of the LSTM network, i.e., the short-term spatiotemporal features.

Further, the historical information v _t-1 and c _t-1 of the short time window are input into the LSTM network.

Furthermore, the historical information [v ₁ , v ₂ , _.

Finally, the long-term spatiotemporal features, the short-term spatiotemporal features and the initial prediction result are respectively input to the output module 104, and the final prediction result is obtained through the output module 104.

In addition, after actual verification in an aquaponics base, the thermal environment prediction model of the present invention achieves excellent results in the multi-step prediction of circulating water temperature. The present invention achieves 15.40%, 13.93%, and 22.15% accuracy improvements in the RMSE indicators in the prediction tasks of 6 hours, 12 hours, and 24 hours, respectively. In particular, the advantages are more pronounced in the long-term prediction of thermal environments. The multi-scale memory structure of the present invention gives the prediction model time-series insight from coarse-grained to fine-grained, and can completely learn the laws in the thermal environment time series data. In addition, the different components used in the present invention can effectively mine the spatiotemporal correlation characteristics in the thermal environment data, so that the prediction model can obtain sufficient spatiotemporal patterns and effectively improve the accuracy of the prediction task.

In the thermal environment prediction model and thermal environment prediction method based on agricultural facilities provided by the present invention, the thermal environment prediction result is obtained through the thermal environment prediction model, wherein the thermal environment prediction model includes: a first memory module, a second memory module, an automatic regression module and output module. The present invention uses the first memory module to perform feature extraction on the time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, wherein the time series information includes: environmental parameters and target parameters of the thermal environment. It can be seen that the thermal environment of the present invention The environmental prediction model effectively obtains the long-term spatial-temporal characteristics of the thermal environment based on the relevant parameters of the thermal environment; the second memory module, based on the long-term spatial-temporal characteristics of the thermal environment, extracts the short-term spatial-temporal characteristics of the thermal environment. The short-term spatiotemporal characteristics of the environment; the autoregressive module is used to determine the initial prediction results based on the historical information of the target parameters; the output module is used to output the final prediction results of the thermal environment based on the long-term spatiotemporal characteristics, short-term spatiotemporal characteristics and the initial prediction results, It can be seen that the present invention fully considers the long-term space-time characteristics and short-term space-time characteristics of the thermal environment, as well as the spatial relationship between the long-term space-time characteristics and the short-term space-time characteristics, so that the present invention can better adapt to the thermal environment prediction of agricultural facilities, and finally achieve accurate Thermal environment prediction for agricultural facilities.

The embodiment of the present invention provides a thermal environment prediction method based on a thermal environment prediction model of agricultural facilities. The thermal environment prediction method described below and the thermal environment prediction model described above refer to each other correspondingly. The specific implementation of the method is shown in the figure 4 shows:

Step 401 , perform feature extraction on the time series information corresponding to the thermal environment to obtain long-term spatiotemporal features of the thermal environment.

The time series information includes: environmental parameters and target parameters of the thermal environment.

Step 402 , based on the long-term spatiotemporal features, extract short-term spatiotemporal features of the thermal environment.

In a specific embodiment, the long-term spatiotemporal features are adjusted, and the short-term spatiotemporal features are determined based on the adjusted long-term spatiotemporal features and historical information.

Step 403: Determine an initial prediction result based on the historical information of the target parameter.

Step 404 , output the final prediction result of the thermal environment based on the long-term spatiotemporal characteristics, the short-term spatiotemporal characteristics and the initial prediction result.

In a specific embodiment, the long-term spatiotemporal feature, the short-term spatiotemporal feature and the initial prediction result are summed to obtain the final prediction result.

FIG. 5 illustrates a schematic diagram of the physical structure of an electronic device. As shown in FIG. 5 , the electronic device may include: a processor (processor) 501, a communication interface (Communications Interface) 502, a memory (memory) 503, and a communication bus 504, The processor 501 , the communication interface 502 , and the memory 503 communicate with each other through the communication bus 504 . The processor 501 can call logic instructions in the memory 503 to execute a thermal environment prediction method, the method includes: performing feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, wherein the time series information includes: Environmental parameters and target parameters of thermal environment; based on long-term spatiotemporal features, extract short-term spatiotemporal features of thermal environment; determine initial prediction results based on historical information of target parameters; output thermal environment based on long-term spatiotemporal features, short-term spatiotemporal features and initial prediction results the final prediction result.

In addition, the above-mentioned logic instructions in the memory 503 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, Read-Only Memory (ROM, Read-Only Memory), Random Access Memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program codes .

In another aspect, the present invention also provides a computer program product, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, when the program instructions are executed by a computer When executing, the computer can execute the thermal environment prediction method provided by the above methods. The method includes: performing feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, wherein the time series information includes: the thermal environment based on the long-term spatiotemporal characteristics, extract the short-term spatiotemporal characteristics of the thermal environment; determine the initial prediction results based on the historical information of the target parameters; output the final thermal environment based on the long-term spatiotemporal characteristics, short-term spatiotemporal characteristics and initial prediction results forecast result.

In another aspect, the present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored, the computer program is implemented by a processor to execute the thermal environment prediction methods provided above, and the method includes: Feature extraction is performed on the time series information corresponding to the thermal environment to obtain the long-term spatiotemporal characteristics of the thermal environment, wherein the time series information includes: environmental parameters and target parameters of the thermal environment; The historical information of the target parameters determines the initial prediction result; based on the long-term spatiotemporal characteristics, short-term spatiotemporal characteristics and the initial prediction result, the final prediction result of the thermal environment is output.

The device embodiments described above are only illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed over multiple network elements. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution in this embodiment. Those of ordinary skill in the art can understand and implement it without creative effort.

From the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, and certainly can also be implemented by hardware. Based on this understanding, the above-mentioned technical solutions can be embodied in the form of software products in essence or the parts that make contributions to the prior art, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic A disc, an optical disc, etc., includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform the methods described in various embodiments or some parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, but not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be The technical solutions described in the foregoing embodiments are modified, or some technical features thereof are equivalently replaced; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. a thermal environment prediction model based on agricultural facilities, is characterized in that, comprises: a first memory module, configured to perform feature extraction on time-series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, where the time-series information includes: environmental parameters and target parameters of the thermal environment; a second memory module, configured to extract short-term spatiotemporal features of the thermal environment based on the long-term spatiotemporal features; The autoregressive module is used to determine the initial prediction result based on the historical information of the target parameter; An output module, configured to output the final prediction result of the thermal environment based on the long-term spatiotemporal features, the short-term spatiotemporal features and the initial prediction result. 2. The thermal environment prediction model based on agricultural facilities according to claim 1, wherein the second memory module comprises: an attention unit and at least two second memory units; The attention unit is used for connecting the first memory module and the second memory module, for adjusting the long-term spatiotemporal features, and transmitting the adjusted long-term spatiotemporal features to the second memory unit. 3. The thermal environment prediction model based on agricultural facilities according to claim 2, wherein the first memory module comprises at least two first memory units; The attention unit is specifically configured to calculate the attention weight corresponding to each of the first memory units based on the long-term space-time feature and the short-term space-time feature after the long-term space-time feature is obtained, and based on the attention force weight, adjust the long-term spatiotemporal feature, and transmit the adjusted long-term spatiotemporal feature to the second memory unit. 4. The thermal environment prediction model based on agricultural facilities according to claim 3, wherein the second memory unit is configured to determine the short-term based on the adjusted long-term spatiotemporal characteristics and the historical information spatiotemporal features. 5. The thermal environment prediction model based on agricultural facilities according to any one of claims 2-4, wherein the second memory module is a recurrent neural network RNN, and the second memory unit comprises: long short-term memory Network LSTM. 6 . The thermal environment prediction model based on agricultural facilities according to claim 3 , wherein the first memory unit comprises: a time-domain convolutional network (TCN). 7 . 7 . The thermal environment prediction model based on agricultural facilities according to claim 1 , wherein the output module is specifically configured to calculate the long-term spatiotemporal features, the short-term spatiotemporal features and the initial prediction results. 8 . and to obtain the final prediction result. 8. A thermal environment prediction method based on the agricultural facility-based thermal environment prediction model according to any one of claims 1-7, characterized in that, comprising: Perform feature extraction on time series information corresponding to the thermal environment to obtain long-term spatiotemporal characteristics of the thermal environment, where the time series information includes: environmental parameters and target parameters of the thermal environment; extracting short-term spatiotemporal features of the thermal environment based on the long-term spatiotemporal features; determining an initial prediction result based on the historical information of the target parameter; Based on the long-term spatiotemporal features, the short-term spatiotemporal features, and the initial prediction result, a final prediction result of the thermal environment is output. 9 . The thermal environment prediction method according to claim 8 , wherein the extracting short-term spatiotemporal features of the thermal environment based on the long-term spatiotemporal features comprises: 10 . The long-term spatiotemporal features are adjusted, and the short-term spatiotemporal features are determined based on the adjusted long-term spatiotemporal features and the historical information. 10 . The thermal environment prediction method according to claim 9 , wherein, based on the long-term spatiotemporal features, the short-term spatiotemporal features and the initial prediction result, outputting the final prediction result of the thermal environment, comprising: 11 . : The long-term spatiotemporal feature, the short-term spatiotemporal feature, and the initial prediction result are summed to obtain the final prediction result.

Images