阅读说明：本技术 一种建筑冷热负荷预测方法、装置、设备及存储介质 (Building cold and heat load prediction method, device, equipment and storage medium ) 是由 曾凯文 刘嘉宁 杜斌 林斌 王可 段秦尉 于 2021-09-10 设计创作，主要内容包括：本发明公开了一种建筑冷热负荷预测方法、装置、设备及存储介质,涉及负荷预测技术领域。所述方法包括获取建筑属性数据集和冷热负荷数据集,并根据预设比例得到训练集和测试集；选取至少三种机器学习模型对所述训练集进行拟合,得到预测结果；根据各所述测试集和各所述预测结果,分析每个所述机器学习模型的预测精度；将预测精度最高的机器学习模型作为最终模型,并将待预测建筑的建筑属性数据输入所述最终模型,得到所述待预测建筑的预测冷热负荷数据。本发明能够解决现有模拟算法建立的模型不够全面,预测出的结果精度不够高的问题。(The invention discloses a method, a device, equipment and a storage medium for predicting cold and hot loads of a building, and relates to the technical field of load prediction. The method comprises the steps of obtaining a building attribute data set and a cold and heat load data set, and obtaining a training set and a testing set according to a preset proportion; selecting at least three machine learning models to fit the training set to obtain a prediction result; analyzing the prediction precision of each machine learning model according to each test set and each prediction result; and taking the machine learning model with the highest prediction precision as a final model, and inputting the building attribute data of the building to be predicted into the final model to obtain the predicted cold and heat load data of the building to be predicted. The invention can solve the problems that the model established by the existing simulation algorithm is not comprehensive enough and the predicted result precision is not high enough.)

1. A method for predicting cold and heat loads of a building is characterized by comprising the following steps:

acquiring a building attribute data set and a cold and heat load data set, and acquiring a training set and a test set according to a preset proportion;

selecting at least three machine learning models to fit the training set to obtain a prediction result;

analyzing the prediction precision of each machine learning model according to each test set and each prediction result;

and taking the machine learning model with the highest prediction precision as a final model, and inputting the building attribute data of the building to be predicted into the final model to obtain the predicted cold and heat load data of the building to be predicted.

2. A method for predicting a cold and thermal load of a building according to claim 1, wherein the building attribute data set comprises a plurality of building attribute data, each of which comprises a relative compactness, a surface area, a wall area, a roof area, a height, an orientation, a light transmission area and a light transmittance of the building.

3. The method for predicting cold and heat loads of a building according to claim 1, wherein the selecting at least three machine learning models to fit the training set to obtain the prediction result comprises:

selecting a Lasso regression model to fit the training set, wherein a specific optimization target formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iAn attribute matrix representing the ith data; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor.

4. The method for predicting cold and heat loads of a building according to claim 1, wherein the selecting at least three machine learning models to fit the training set to obtain the prediction result comprises:

selecting a Bagging algorithm model to fit the training set, randomly extracting a sample from the training set of m samples and putting the sample into a sampling set, then putting the sample back into the sampling set, and obtaining the sampling set containing n samples after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMinimizing the prediction error; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

5. A building cold and heat load prediction device, comprising:

the data acquisition module is used for acquiring the building attribute data set and the cold and heat load data set and obtaining a training set and a test set according to a preset proportion;

the fitting prediction module is used for selecting at least three machine learning models to fit the training set to obtain a prediction result;

a model analysis module for analyzing the prediction accuracy of each machine learning model according to each of the test sets and each of the prediction results;

and the model screening module is used for taking the machine learning model with the highest prediction precision as a final model and inputting the building attribute data of the building to be predicted into the final model to obtain the predicted cold and heat load data of the building to be predicted.

6. A cold and thermal load forecasting apparatus according to claim 5, wherein the building attribute data set comprises a plurality of building attribute data, each building attribute data comprising a relative compactness, surface area, wall area, roof area, height, orientation, light transmission area and light transmission of the building.

7. The building cold and heat load prediction device according to claim 5, wherein the fitting prediction module is configured to select a Lasso regression model to fit the training set, and a specific optimization objective formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iAn attribute matrix representing the ith data; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor.

8. The building cold and heat load prediction device according to claim 5, wherein the fitting prediction module is configured to select a Bagging algorithm model to fit the training set, randomly extract a sample from the training set of m samples and put the sample into the sampling set, then put the sample back into the sampling set, and obtain a sampling set including n samples after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMinimizing the prediction error; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

9. A computer terminal device, comprising:

one or more processors;

a memory coupled to the processor for storing one or more programs;

when executed by the one or more processors, cause the one or more processors to implement a building cold and thermal load prediction method as claimed in any one of claims 1 to 4.

10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements a building cold and heat load prediction method according to any one of claims 1 to 4.

Technical Field

The invention relates to the technical field of load prediction, in particular to a method, a device, equipment and a storage medium for predicting cold and hot loads of a building.

Background

The commercial buildings are typical units in public buildings, have the characteristic of high power consumption density in unit area, and with the rapid development of the economic society, the commercial building area is rapidly expanded, so that the load level of the commercial buildings is continuously improved. The method can accurately predict the cold and heat loads of the commercial building, can provide reliable data for power grid dispatching, and has a prominent guiding function on the operation of power equipment, the economic dispatching of the power grid, power marketing and other works.

At present, methods for predicting cold and hot loads of buildings at home and abroad are mainly classified into energy consumption software simulation prediction, a traditional prediction method based on statistical regression analysis and an artificial intelligence prediction algorithm based on machine learning. The method for utilizing the energy consumption simulation software is generally based on software platforms such as EnergyPlus, TRANSYS, DOE-2 and the like, a detailed building model is built, relevant design parameters are given, and relevant load predicted values are obtained through computer simulation. However, the method for predicting the load by the energy consumption simulation software has higher requirements on operators and basic data, and is often used as an auxiliary means rather than a main method for predicting the load. The traditional prediction method based on statistical regression analysis is simple in principle and wide in application, mainly establishes a functional relation between loads and influence factors through a large amount of basic data, but a model established by the algorithm is not comprehensive enough, the predicted result is not high enough in precision, and a complex nonlinear model is difficult to accurately describe.

Disclosure of Invention

The invention aims to provide a method, a device, equipment and a storage medium for predicting cold and hot loads of a building, so as to solve the problems that a model established by the existing simulation algorithm is not comprehensive enough and the predicted result precision is not high enough.

In order to achieve the above object, an embodiment of the present invention provides a method for predicting a cold and heat load of a building, including:

acquiring a building attribute data set and a cold and heat load data set, and acquiring a training set and a test set according to a preset proportion;

selecting at least three machine learning models to fit the training set to obtain a prediction result;

analyzing the prediction precision of each machine learning model according to each test set and each prediction result;

and taking the machine learning model with the highest prediction precision as a final model, and inputting the building attribute data of the building to be predicted into the final model to obtain the predicted cold and heat load data of the building to be predicted.

Preferably, the building attribute data set comprises a number of building attribute data, each building attribute data comprising a relative compactness, a surface area, a wall area, a roof area, a height, an orientation, a light transmission area and a light transmission of the building.

Preferably, the selecting at least three machine learning models to fit the training set to obtain a prediction result includes:

selecting a Lasso regression model to fit the training set, wherein a specific optimization target formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iGenus representing ith dataA property matrix; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor.

Preferably, the selecting at least three machine learning models to fit the training set to obtain a prediction result includes:

selecting a Bagging algorithm model to fit the training set, randomly extracting a sample from the training set of m samples and putting the sample into a sampling set, then putting the sample back into the sampling set, and obtaining the sampling set containing n samples after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMinimizing the prediction error; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

The embodiment of the invention also provides a device for predicting the cold and heat load of the building, which comprises:

the data acquisition module is used for acquiring the building attribute data set and the cold and heat load data set and obtaining a training set and a test set according to a preset proportion;

the fitting prediction module is used for selecting at least three machine learning models to fit the training set to obtain a prediction result;

a model analysis module for analyzing the prediction accuracy of each machine learning model according to each of the test sets and each of the prediction results;

and the model screening module is used for taking the machine learning model with the highest prediction precision as a final model and inputting the building attribute data of the building to be predicted into the final model to obtain the predicted cold and heat load data of the building to be predicted.

Preferably, the building attribute data set comprises a number of building attribute data, each building attribute data comprising a relative compactness, a surface area, a wall area, a roof area, a height, an orientation, a light transmission area and a light transmission of the building.

Preferably, the fitting prediction module is configured to select a Lasso regression model to fit the training set, and a specific optimization target formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iAn attribute matrix representing the ith data; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor.

Preferably, the fitting prediction module is configured to select a Bagging algorithm model to fit the training set, randomly extract a sample from the training set of m samples and place the sample into the sampling set, then place the sample back into the sampling set, and obtain a sampling set including n samples after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMake the prediction wrongThe difference is minimal; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

The embodiment of the invention also provides computer terminal equipment which comprises one or more processors and a memory. A memory coupled to the processor for storing one or more programs; when executed by the one or more processors, cause the one or more processors to implement a method for building cold and heat load prediction as described in any of the embodiments above.

Embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the method for predicting the cold and heat load of a building according to any of the above embodiments.

Compared with the prior art, the invention has the following beneficial effects:

the invention discloses a building cold and heat load prediction method, which comprises the steps of obtaining a building attribute data set and a cold and heat load data set, and obtaining a training set and a test set according to a preset proportion; selecting at least three machine learning models to fit the training set to obtain a prediction result; analyzing the prediction precision of the machine learning model according to the test set and the prediction result; according to the prediction precision, selecting the model with the highest precision in the machine learning model as a final model; and inputting the building attribute data to be predicted into the final model to obtain predicted cold and heat load data. The method integrates various machine learning models to fit the building attribute data set and the cold and heat load data set, and preferably selects the most suitable machine learning model to input the building attribute data to be predicted into the final model to obtain predicted cold and heat load data, so that the problems that the model established by the conventional simulation algorithm is not comprehensive enough and the predicted result is not high in precision can be solved.

Drawings

In order to more clearly illustrate the technical solution of the present invention, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a schematic flow chart of a method for predicting a cold and heat load of a building according to an embodiment of the present invention;

fig. 2 is a schematic structural diagram of a building cold and heat load prediction device according to an embodiment of the present invention;

fig. 3 is a schematic structural diagram of a computer terminal device according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

It should be understood that the step numbers used herein are for convenience of description only and are not used as limitations on the order in which the steps are performed.

It is to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.

The terms "comprises" and "comprising" indicate the presence of the described features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.

The term "and/or" refers to and includes any and all possible combinations of one or more of the associated listed items.

Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for predicting a building cold and heat load according to an embodiment of the present invention. In this embodiment, the method for predicting the cold and heat load of the building includes the following steps:

s110, acquiring a building attribute data set and a cold and heat load data set, and obtaining a training set and a test set according to a preset proportion;

s120, selecting at least three machine learning models to fit the training set to obtain a prediction result;

in one embodiment, fitting is performed on the training set, and the machine learning model used comprises: a ridge regression model, a random forest model, a Bagging algorithm model, a limit tree model and a K-nearest neighbor model.

S130, analyzing the prediction precision of each machine learning model according to each test set and each prediction result;

and S140, taking the machine learning model with the highest prediction precision as a final model, and inputting the building attribute data of the building to be predicted into the final model to obtain the predicted cold and heat load data of the building to be predicted.

In an embodiment of the present invention, the building attribute data set includes a plurality of building attribute data, each of which includes a relative compactness, a surface area, a wall area, a roof area, a height, an orientation, a light transmission area, and a light transmittance of the building.

In this embodiment of the present invention, step S120, selecting at least three machine learning models to fit the training set, so as to obtain a prediction result, includes:

selecting a Lasso regression model to fit the training set, wherein a specific optimization target formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iAn attribute matrix representing the ith data; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor.

In this embodiment of the present invention, step S120, selecting at least three machine learning models to fit the training set, so as to obtain a prediction result, includes:

selecting a Bagging algorithm model to fit the training set, randomly extracting a sample from the training set of m samples and putting the sample into a sampling set, then putting the sample back into the sampling set, and obtaining the sampling set containing n samples after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMinimizing the prediction error; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

In a specific embodiment, the self attribute and the cold and heat load data of a certain commercial building are obtained, and a training set and a test set are constructed; selecting 6 typical models of ridge regression, random forest, Bagging, extreme tree and K-nearest neighbor in machine learning to fit a training set, and testing the prediction precision of the model on a test set by using a cross validation method; selecting three models with the best effect by using a cross validation method, and fusing the selected models by using a model fusion method to obtain a final prediction model; and finally, applying a final prediction model to predict the cold and heat load of the commercial building.

In this embodiment, the obtained attributes of the commercial building itself include relative compactness, surface area, wall area, roof area, height, orientation, light transmission area, and light transmittance. The obtained commercial building data is divided into training sets and tests according to the distribution proportion of 80% training sets and 20% testing sets.

In this embodiment, for the training set, fitting is performed by using a Lasso regression model, and a specific optimization target formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iAn attribute matrix representing the ith data; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor. And (3) calculating to obtain a coefficient matrix omega by using a random gradient descent method with the minimum of the formula (1) as a target. | ω | non-woven gas in the above formula₁It can also be replaced by a 2 norm of the coefficient matrix: | ω | non-calculation₂. Norm, which is a function with the concept of "length". In the fields of linear algebra, functional analysis and related mathematics, a norm is a function that assigns all vectors in vector space a positive length or magnitude that is non-zero. The half-norm may instead assign a zero length to a non-zero vector. The two-norm is the square root value of the maximum eigenroot of the product of the transposed conjugate matrix of matrix a and matrix a, and is the linear distance between two vector matrices in space. Similar to finding the straight-line distance between two points on the chessboard.

In this embodiment, for the training set, a Bagging algorithm is used for fitting, a sample is randomly extracted from the training set of m samples and placed into the sampling set, then the sample is placed back into the sampling set, and a sampling set containing n samples is obtained after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMinimizing the prediction error; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

In this embodiment, for the training set, a random forest algorithm is used for fitting. The random forest selects a decision tree as a base learner based on a Bagging algorithm. That is, for each node of each base decision tree, a subset containing a partial feature set is randomly selected from the feature set of the node, and then an optimal attribute is selected from the subset for partitioning. The remaining features are the same as fitting using the Bagging algorithm.

In this embodiment, for the training set, a limit tree model is used for fitting. The extreme tree model is based on a random forest model, and in the process of training the base learner, a new sample set is not extracted and constructed randomly from an original training set any more, but the base learner is generated on the basis of all training samples.

In a specific embodiment, in step S140, the highest accuracy in the machine learning model is selected as the final model according to the prediction accuracy.

For each model, the training data set is divided into k mutually exclusive subsets of similar size, i.e., D ═ D₁∪D₂∪...∪D_k. Where D represents the entire training data set. Training is performed on the union of k-1 subsets, and validation is performed on the remaining set. Each subset is used for verification once, and the cycle is repeated for k times, and the average value and the standard deviation of the prediction precision are obtained. And selecting a model with a larger average value and a smaller standard deviation, wherein the specific result is shown in table 1, and selecting Bagging, random forest and limit tree as a base model for model fusion according to the cross validation result.

Table 1 Cross-validation results of the models

And then, fusing the selected models by adopting a learning model fusion technology. Assuming the selected J models m₁，m₂，...m_J. Firstly, the training characteristics in the training set are input into the selected model m_jThen, a set of predicted values is obtainedWhereinThe representation model j predicts the nth data in the data set. And taking the generated predicted value as a characteristic value of the fusion model, taking the actual load number of the original data set as a true value, and fitting by adopting a multi-response linear regression algorithm to obtain a final prediction model.

In the embodiment of the invention, the building attribute data to be predicted is input into the final model to obtain the predicted cold and heat load data. The calculation effects of this example are given below

Table 2 shows the predicted results of various models, and the effect of the models after fusion. From the results, it can be seen that the three models have higher performance on the task of predicting the building load, especially on the prediction of the heat load, but the prediction of the cold load still has a more obvious error. Therefore, the prediction results of the three models are subjected to weighted fusion in a model fusion mode, and the load prediction accuracy and the generalization performance of the models are improved. It can be seen from the results that, compared to Bagging, the generation of the tree bifurcation attributes of the random forest and the limit tree is completely random for all data, and has higher randomness, so that better performance is obtained. After model fusion is added, the prediction precision of the heat load and the cold load is improved, which shows that the model fusion gives higher weight to the object with more accurate prediction of each model on the basis of the existing model, which is equivalent to the retention of the advantages of each model, and further improves the performance of the model.

Table 2 results of examples

The invention discloses a building cold and heat load prediction method, which comprises the steps of obtaining a building attribute data set and a cold and heat load data set, and obtaining a training set and a test set according to a preset proportion; selecting at least three machine learning models to fit the training set to obtain a prediction result; analyzing the prediction precision of the machine learning model according to the test set and the prediction result; according to the prediction precision, selecting the model with the highest precision in the machine learning model as a final model; and inputting the building attribute data to be predicted into the final model to obtain predicted cold and heat load data. The invention can solve the problems that the model established by the existing simulation algorithm is not comprehensive enough and the predicted result precision is not high enough.

Referring to fig. 1, fig. 1 is a schematic flow chart illustrating a method for predicting a building cold and heat load according to an embodiment of the present invention. In this embodiment, the building cold and heat load prediction device includes:

the data acquisition module 210 is configured to acquire a building attribute data set and a cold and heat load data set, and obtain a training set and a test set according to a preset ratio;

a fitting prediction module 220, configured to select at least three machine learning models to fit the training set, so as to obtain a prediction result;

a model analysis module 230, configured to analyze the prediction accuracy of the machine learning model according to the test set and the prediction result;

the model screening module 240 is configured to select, according to the prediction accuracy, a final model with the highest accuracy in the machine learning model;

and the target prediction module 250 is used for inputting the building attribute data to be predicted into the final model to obtain predicted cold and heat load data.

In embodiments of the invention, the building attribute data set comprises relative compactness, surface area, wall area, roof area, height, orientation, light transmission area and light transmission of the building.

In the embodiment of the present invention, the fitting prediction module is configured to select a Lasso regression model to fit the training set, and a specific optimization target formula is as follows:

wherein J (omega) represents an objective function taking omega as an optimization variable, and variable m represents the number of data in a training set; y is_iRepresenting the actual load number of the ith data; x_iAn attribute matrix representing the ith data; ω represents an attribute coefficient matrix; | ω | non-calculation₁A 1 norm representing a matrix of attribute coefficients; λ is a penalty factor.

In the embodiment of the invention, the fitting prediction module is used for selecting a Bagging algorithm model to fit the training set, randomly extracting a sample from the training set of m samples and putting the sample into a sampling set, then putting the sample back into the sampling set, and obtaining the sampling set containing n samples after n times of sampling; sampling T sampling sets containing n training samples; training based on each sampling set to obtain a base learner h_t(ii) a Aggregating the results of the T base learners based on a simple voting method, wherein the concrete formula is as follows:

wherein l represents a loss function, i.e. the difference between the predicted value and the true value;weight w representing selection of each base learner_tMinimizing the prediction error; h is_t(x_i) Presentation basis learning machine h_tFor sample x_iAnd (4) predicting.

For specific limitations of the building cold and heat load prediction device, reference may be made to the above limitations of the building cold and heat load prediction method, which will not be described herein again. The modules in the building cold and heat load prediction device can be wholly or partially realized by software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.

Referring to fig. 3, an embodiment of the invention provides a computer terminal device, which includes one or more processors and a memory. The memory is coupled to the processor for storing one or more programs which, when executed by the one or more processors, cause the one or more processors to implement the building cold and heat load prediction method as in any one of the embodiments described above.

The processor is used for controlling the overall operation of the computer terminal equipment so as to complete all or part of the steps of the building cold and heat load prediction method. The memory is used to store various types of data to support the operation at the computer terminal device, which data may include, for example, instructions for any application or method operating on the computer terminal device, as well as application-related data. The Memory may be implemented by any type of volatile or non-volatile Memory device or combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read-Only Memory (EPROM), Programmable Read-Only Memory (PROM), Read-Only Memory (ROM), magnetic Memory, flash Memory, magnetic disk, or optical disk.

In an exemplary embodiment, the computer terminal Device may be implemented by one or more Application Specific 1 integrated circuits (AS 1C), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a controller, a microcontroller, a microprocessor or other electronic components, for performing the above-mentioned building cold and heat load prediction method, and achieving technical effects consistent with the above-mentioned methods.

In another exemplary embodiment, a computer readable storage medium is also provided, which comprises program instructions, which when executed by a processor, implement the steps of the building cold and heat load prediction method in any one of the above embodiments. For example, the computer readable storage medium may be the memory including program instructions executable by a processor of a computer terminal device to perform the method for predicting the cold and heat load of a building as described above, and achieve the same technical effects as the method described above.

While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.