CN117829623A - Airport terminal cold load prediction method and system based on RF-IGOA-LSTM model - Google Patents

Abstract

The invention discloses a cold load prediction method and a system for airport terminal based on an RF-IGOA-LSTM model, which are used for preprocessing data; extracting the characteristic value of the input parameter by using a random forest algorithm; introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm; optimizing the super parameters in the LSTM algorithm by using the improved locust optimization algorithm; and predicting the cold load of the airport terminal by using the optimized IGOA-LSTM model. The method can effectively screen the input characteristics of the dynamic cold load model of the airport terminal, and has the performance advantages of high prediction precision, high running speed, strong robustness and strong generalization capability in the related building application, thereby realizing the effective prediction of seasonal dynamic change cold load.

Description

Airport terminal cold load prediction method and system based on RF-IGOA-LSTM model

Technical Field

The invention belongs to the technical field of cold load prediction of airport air conditioning systems, and particularly relates to a cold load prediction method and system of an airport terminal based on an RF-IGOA-LSTM model.

Background

In recent years, the development of air transportation is rapid and gradually becomes an important component of the economic development of China, and more airports continuously expand the airport terminal, the runway and other measures to relieve the pressure of air transportation. With the continuous expansion of the airport scale, the airport energy consumption is continuously increased, the problems of high energy consumption density, low energy utilization rate and the like are endlessly layered, and meanwhile, the airport energy consumption ratio of the terminal building is maximum, so that huge energy saving potential exists. Meanwhile, the airport terminal building is used as a main functional area in airport buildings, and has the characteristics of various basic functional partitions of the building, complex running equipment and related systems, dynamic change of environmental parameters along with outdoor weather parameters and indoor personnel and the like, so that the accurate cold load prediction of the terminal building is particularly critical by an effective and reliable method.

With the progress and development of computer science, more and more meta-heuristic algorithms are widely applied to various machine learning models after effective improvement, and show excellent prediction performance and results in data-driven cold load prediction. Therefore, reasonable cold load prediction can improve the energy utilization efficiency and reduce the total energy waste while ensuring the normal operation of an airport, and likewise, the establishment of a related model has certain reference significance and reference price for development planning of the airport and energy-saving optimization and solving the expansibility of an energy system optimization scheduling of a regional cooling system.

Disclosure of Invention

The technical problem to be solved by the invention is to provide a cold load prediction method and a system for an airport terminal based on an RF-IGOA-LSTM model aiming at the defects in the prior art, which are used for solving the technical problem of cold load unstable prediction under the indoor and outdoor dynamic parameter change of the airport terminal.

The invention adopts the following technical scheme:

an airport terminal cold load prediction method based on an RF-IGOA-LSTM model comprises the following steps:

s1, dividing airport data according to seasons according to the characteristic that the cold load of an airport terminal changes along with weather, and realizing data preprocessing;

s2, carrying out feature extraction on the data obtained in the step S1 by using a random forest;

s3, introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm;

s4, optimizing the super parameters in the LSTM algorithm by using the locust optimization algorithm improved in the step S3 to obtain an IGOA-LSTM prediction model;

s5, inputting the data obtained in the step S2 into the IGOA-LSTM model optimized in the step S4, and predicting the cold load of the airport terminal.

Preferably, in the data preprocessing process, linear interpolation is used to fill in the data, and then normalization processing is performed on the max-min.

Preferably, the data obtained in the step S1 is subjected to feature extraction by using a random forest, and parameters with weight values larger than 0.8 are selected.

More preferably, the feature extraction of the data using random forests is specifically:

generating n training subsets by using a bootstrap sampling method according to an original training set, wherein each training set has m samples in total;

model training is carried out on the n training subsets, and n decision tree models are generated;

for the feature attribute of a single decision tree model, selecting the strongest segmentation capability feature from all features by using a minimum variance criterion to perform node segmentation;

under the condition of no pruning, integrating each tree into a random forest after the maximum growth, and then calculating the influence of the data outside the bag on the model by using a mean square error;

voting on all decision trees, i.e. the importance of the input variables is represented by a reduction of MSE, and the mean square residual sequence is obtained by using the out-of-bag data MSE1, MSE 2.

Preferably, the position update formula of the locust individual introducing the cauchy-gaussian factor to improve the traditional locust optimization algorithm is as follows:

wherein,for the random position of the kth locust on the d dimension, gaussian is Gaussian operator, +. >For the optimal position of the t generation in the d-th dimension,>is a random position of the t generation in the d dimension.

Preferably, the improved locust optimization algorithm is utilized to optimize the super parameter learning rate and the number of hidden layer nodes in the LSTM algorithm, and a group of parameters which minimize the loss function is found to obtain an IGOA-LSTM prediction model, which is specifically as follows:

where N 'is the number of samples in the test dataset, y' _i Is the actual value of the test data set,is the corresponding predicted output value.

Preferably, the prediction of the cold load of the airport terminal is specifically:

according to the result of random forest algorithm operation, selecting the corresponding characteristic as an input variable of an IGOA-LSTM prediction model, and taking the cold load of an airport terminal as an output variable of the IGOA-LSTM prediction model; training an IGOA-LSTM prediction model by means of actual airport building data to obtain accurate airport terminal cold load prediction values.

In a second aspect, an embodiment of the present invention provides a system for predicting a cooling load of an airport terminal based on an RF-IGOA-LSTM model, including:

the data module divides airport data according to seasons according to the characteristic that the cold load of the airport terminal changes along with weather, so as to realize data preprocessing;

The extraction module is used for extracting characteristics of the obtained data by using a random forest;

the improvement module is used for introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm;

the optimization module optimizes the super parameters in the LSTM algorithm by utilizing the improved locust optimization algorithm to obtain an IGOA-LSTM prediction model;

and the prediction module inputs the data into the optimized IGOA-LSTM model to realize the prediction of the cold load of the airport terminal.

In a third aspect, a chip includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the method for predicting the cold load of an airport terminal based on the RF-IGOA-LSTM model.

In a fourth aspect, an embodiment of the present invention provides an electronic device, including a computer program, where the computer program when executed by the electronic device implements the steps of the method for predicting a cold load of an airport terminal based on the RF-IGOA-LSTM model.

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

according to the method, the cold load input characteristics of the large airport terminal are effectively screened, the interpolation and the maximum-minimum normalization preprocessing of the cold load data of the existing airport terminal are carried out, the data are divided according to the characteristics of the cold load of the large airport terminal along with the climate change, the data are calculated and analyzed according to the seasons, the importance of the input characteristics of the model is ordered by means of random forests, obvious influence characteristics are selected, the LSTM network structure and the parameters are optimized by means of an improved locust optimization algorithm, the cold load prediction model is trained and verified, the structure of the cold load prediction model is finally established, the model redundancy and the complexity of the characteristic model are greatly reduced through the characteristic selection of the random forests and the introduction of the improved optimization algorithm, and the operation efficiency of the prediction model is improved. Experiments prove that the device operation and energy consumption conditions of the cooling system of the large airport terminal can be reflected timely and accurately in practical application, and effective data reference is provided for the prediction of the cooling load of the large airport terminal.

Further, reasonable data preprocessing can greatly improve the efficiency of data processing and modeling and the precision of a prediction result, and the rationality of data is the basis for guaranteeing effective training of a model, so that firstly linear interpolation processing of numerical value estimation is carried out according to specific gravity on two data which are close to each other about an interpolation point in a data sequence, and secondly, the maximum-minimum processing of a multi-unit data set is carried out, so that the characteristics have the same measurement scale, the model precision can be improved, and meanwhile, the convergence speed of the model is improved.

Further, considering that the relevance of part of input data and output data is low and the coupling phenomenon exists in the data, carrying out random forest characteristic analysis on the data, and selecting parameters with weight values larger than 0.8 as model input.

Further, there are many disturbance factors that affect the cooling load, and there is a ubiquitous interaction between data sets. Conventional data analysis methods have difficulty in avoiding interactions between independent variables. As one of the emerging supervised machine learning algorithms, random Forests (RF) have less demand for data sets, run stably, and have no overfitting and collineation problems. The nonlinear, collinear and interactive data can be effectively analyzed, so that the weight of the input features of the training model is calculated by using a random forest, the importance of the model influence degree is ordered, obvious influence features can be selected, the dimensionality of the model input parameters is reduced, and the prediction precision of the cold load prediction model is finally improved.

Further, at the later stage of the HHO algorithm iteration, multiple individuals are easily aggregated, resulting in increased risk of being trapped in a local optimum. To prevent the algorithm from falling into a dead state, a cauchy-gaussian variation strategy is introduced. And when each iteration is finished, selecting an individual with the optimal fitness to carry out cauchy-Gaussian variation, and if the position after variation is better than the current position, selecting an individual with a better position to enter the next iteration. The Cauchy distribution has strong disturbance capability, can improve the global searching capability of the algorithm, and the Gaussian distribution can enhance the local development capability of the algorithm. And the cauchy disturbance is mainly carried out at the initial stage of iteration, so that the HHO is disturbed in a large range, and the global searching capability of the algorithm is improved. When the iterative later stage is entered, gaussian disturbance is mainly performed at the moment, so that HHO is searched near the hunting position, the local development capability of an algorithm is enhanced, and the convergence accuracy is improved. When the early-stage adults are subjected to global optimization, a Cauchy operator is introduced to increase the searching step length of the adults, and the global optimal solution is found out rapidly with higher probability.

Furthermore, gaussian operators are introduced during the later stage of larva local development, so that the local searching capacity is enhanced, and the convergence rate is improved.

Furthermore, the LSTM has the limitation of long-time sequence under the multi-dimensional problem on the prediction model, so that the time span of the LSTM is larger in the model calculation process, the network structure is deeper and more complex, and therefore, the complexity of data processing can be simplified to a great extent and the model processing efficiency can be improved by optimizing the super-parameter learning rate of key factors and the number of hidden layer nodes in the LSTM. And optimizing the super-parameter learning rate and the number of hidden layer nodes in the LSTM algorithm by using an improved locust optimization algorithm (IGOA), and finding a group of parameters which minimize the loss function, thereby obtaining an optimal LSTM prediction model.

It will be appreciated that the advantages of the second aspect may be found in the relevant description of the first aspect, and will not be described in detail herein.

In summary, the cold load prediction method and system for the airport terminal based on the RF-IGOA-LSTM model analyze the energy characteristics and the cold load change rules of large public buildings with similar data sets and building characteristics, firstly, seasonal energy use condition distinction is carried out on the existing historical data sets, then, interpolation-normalization preprocessing of the data sets and characteristic value extraction of random forests are carried out on relevant influence variables of the cold load of the airport terminal, variables with high relevance are selected as model input, a Kexi-Gaussian factor strategy is introduced to improve the traditional GOA algorithm to optimize LSTM network structure and parameters, training and establishment of the prediction model are completed, the dimension of the data sets under the influence of the coupling relation is reduced, the operation speed and accuracy of modeling are improved, and the practical engineering problem of cold load prediction of similar buildings such as the airport terminal is solved.

The technical scheme of the invention is further described in detail through the drawings and the embodiments.

Drawings

FIG. 1 is a flow chart of the method of the present invention;

FIG. 2 is a flowchart of the algorithm setup of the IGOA-LSTM of the present invention;

FIG. 3 is an iterative optimization performance diagram of the IGOA algorithm of the invention;

FIG. 4 is a graph of the fitting effect of the model of the present invention;

FIG. 5 is a comparison of feature extraction and non-feature extraction of the model of the present invention;

FIG. 6 is a schematic diagram of a computer device according to an embodiment of the present invention;

fig. 7 is a block diagram of a chip according to an embodiment of the present invention.

Detailed Description

The following description of the embodiments of the present invention will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

In the description of the present invention, it will be understood that the terms "comprises" and "comprising," when used in this specification, specify the presence of stated 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.

It is also to be understood that the terminology used in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this specification 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.

It should be further understood that the term "and/or" as used in the present specification and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes such combinations, e.g., a and/or B, may represent: a exists alone, A and B exist together, and B exists alone. In the present invention, the character "/" generally indicates that the front and rear related objects are an or relationship.

It should be understood that although the terms first, second, third, etc. may be used to describe the preset ranges, etc. in the embodiments of the present invention, these preset ranges should not be limited to these terms. These terms are only used to distinguish one preset range from another. For example, a first preset range may also be referred to as a second preset range, and similarly, a second preset range may also be referred to as a first preset range without departing from the scope of embodiments of the present invention.

Depending on the context, the word "if" as used herein may be interpreted as "at … …" or "at … …" or "in response to a determination" or "in response to detection". Similarly, the phrase "if determined" or "if detected (stated condition or event)" may be interpreted as "when determined" or "in response to determination" or "when detected (stated condition or event)" or "in response to detection (stated condition or event), depending on the context.

Various structural schematic diagrams according to the disclosed embodiments of the present invention are shown in the accompanying drawings. The figures are not drawn to scale, wherein certain details are exaggerated for clarity of presentation and may have been omitted. The shapes of the various regions, layers and their relative sizes, positional relationships shown in the drawings are merely exemplary, may in practice deviate due to manufacturing tolerances or technical limitations, and one skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions as actually required.

The invention provides a cold load prediction method of an airport terminal based on an RF-IGOA-LSTM model, which can effectively screen model input features by combining an optimization long-short-term memory neural network (IGOA-LSTM) with a Random Forest (RF) through improving a locust algorithm, has the advantages of high prediction precision, high running speed, strong robustness and strong generalization capability, and can effectively predict the cold load of the airport terminal.

Referring to fig. 1, the method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model of the present invention includes the following steps:

s1, preprocessing data, and dividing the data into seasons according to the characteristic that the cold load of a large airport terminal changes along with weather;

the existing airport data set is subjected to interpolation to process missing and abnormal values, and meanwhile, the pretreatment is carried out by means of maximum-minimum normalization. Meanwhile, the existing data set is divided according to seasons according to the characteristic that the cold load of the large airport terminal changes along with the climate.

Missing value processing: the data is padded using linear interpolation. It carries out numerical value estimation according to the proportion of two data which are adjacent to the left and right of the point needing interpolation in the data sequence.

Data normalization: the data normalization can enable the features to have the same measurement scale, so that the model precision can be improved, and meanwhile, the convergence rate of the model can also be improved. The invention uses normalization processing to the maximum-minimum, the processed data is between [0,1 ]; the specific formula is as follows:

wherein, x and y respectively correspond to the data before and after normalization, x _max ，x _min Respectively corresponding to the maximum value and the minimum value of the data before processing.

S2, extracting characteristic values of input parameters by adopting a random forest algorithm (RF);

and (2) calculating the weight of the input features in the step (S1) by using a random forest, and sequencing the importance of the model influence degree, so that the significant influence features can be selected, the dimension of the model input parameters is reduced, and finally, the prediction precision of the cold load prediction model is improved.

The disturbance factors affecting the cooling load are numerous and there is an interaction between them. Conventional data analysis methods have difficulty in avoiding interactions between the arguments. While as one of the emerging supervised machine learning algorithms, random Forests (RF) have less demand for data sets and run stably without overfitting and collineation problems. The method can effectively analyze the data of nonlinearity, collineation and interaction, thereby well extracting factors with great influence on the cold load.

RF consists of decision trees and a bagging framework, where each tree is independent of the others. And randomly selecting a plurality of samples from the original training sample set by using a bootstrap sampling method, generating new training sample subsets with differences, and then establishing a decision tree based on the new training sample sets. When sampling a subset of samples, the unextracted samples are out-of-bag data.

The voting results of the decision tree classification in RF are used to determine the final characteristic impact coefficients. K training subsamples are generated by randomly selecting s samples with substitutions from the complete training set in the decision tree. The OOB data subset is used to form a test set of the model.

For k training subsamples, k independent decision trees are constructed.

Each decision tree grows to a maximum size without any pruning; the minimum variance criterion is used for a node segmentation algorithm, and each decision tree selects the strongest classification characteristic node for segmentation.

Constructing a random forest set classifier by aggregating k decision trees, and classifying corresponding OOB data; the coefficient of influence of the test input feature is determined by calculating a Root Mean Square Error (RMSE).

The effect of each decision tree on the OOB data features can be discerned from its contribution to the RMSE sequence [ RMSE1, RMSE2, …, RMSEk ], and feature importance coefficients are obtained by aggregating votes among all decision trees.

The use of RF extraction features mainly includes the following stages:

(1) From the original training set, n training subsets are generated using a bootstrap sampling method, with a total of m samples per training set.

(2) Model training is carried out on the n training subsets, and n decision tree models are generated.

(3) For the feature attributes of a single decision tree model, selecting the strongest segmentation capability feature from all features by using a minimum variance criterion for node segmentation, wherein the method comprises the following steps:

wherein,i is the optimal segmentation variable, s is the embedded sample dimension, X _s And X' _s The values and average values of the variables are represented respectively, and the total feature number in the feature space of the q decision tree.

(4) Without pruning, each tree is integrated into a random forest after maximum growth, and then the effect of the out-of-bag data on the model is calculated using mean square error (RMSE) as follows:

wherein,and y _i The actual and predicted airport cooling loads, respectively, n is the total number of samples for a given test set and i is the number of samples for the test set.

(5) Voting is performed on all decision trees, i.e. the importance of the input variables is represented by a reduction of MSE, and the mean square residual sequence is obtained by using the out-of-bag data MSE1, MSE2, …, MSEk. For input variables, the influence coefficients are as follows:

wherein Sci is the influence coefficient of the input variable, S _e Is the standard error of n decision trees.

The commonly used input data of the air conditioner cold load prediction comprises a cold load at the previous moment, an outdoor temperature, carbon dioxide, outdoor humidity, wind speed, wind direction, a solar radiation value at the previous moment, time, a solar radiation value at the current moment and the number of indoor people, and the air conditioner cold load at the current moment is used as output.

And taking the fact that the association degree of part of input data and output data in the data is low and the coupling phenomenon exists into consideration, carrying out random forest characteristic analysis on the data, and selecting parameters with weight values larger than 0.8 as model input. The index results of the respective influencing factors after RF analysis are shown in table 1.

Table 1 weighting results of input parameters

From the results of the analysis in table 1, wind direction variables with less influence can be eliminated.

S3, introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm (GOA);

the locust group is divided into adults responsible for global search and larvae developed in a specific adjacent area, and the core formula of the GOA for updating the locust position is as follows:

wherein x is _o+1,d For the coordinate value of the ith locust individual on the d dimension of the position vector, c is a dynamic attenuation coefficient related to the maximum iteration number and capable of affecting the global optimizing or local searching capability of the locust, ub _d And lb (L) _d Respectively the upper and lower threshold limits on the d-th dimension of the individual locusts, S is the interaction force function of the locusts i affected by the locusts j, d _ij T 'is the absolute distance between locust i and locust j' _d The optimal position obtained before the locust population; the convergence rate is increased, thus improving the optimization performance of GOA by introducing a cauchy-gaussian factor.

Furthermore, the core formula with the locust position can know that the locust individual is easy to fall into local optimum when searching the optimum position, so that the optimization performance of the algorithm can be improved to a great extent by introducing the cauchy-Gaussian mixture variation in the position updating formula. When the early-stage adults are subjected to global optimization, a cauchy operator is introduced to increase the searching step length of the adults, and a global optimal solution is quickly found with higher probability, wherein the position updating formula of the locust individuals is as follows:

wherein X is _new (t+1) is the position of the ith locust at the t+1 th generation, X _i (t) is the position of the ith locust in the t generation, X _best (t) is the optimal position of the t generation, X _K (t) is the random position of the t th generation and Cauchy is the Cauchy operator.

Furthermore, gaussian operators are introduced during the later stage of larva local development, so that the local searching capability is enhanced, the convergence rate is improved, and the position updating formula of the locust individual is as follows:

wherein,is the random position of the kth locust in the d dimension, and Gaussian is a Gaussian operator.

S4, optimizing the super parameters in the LSTM algorithm by using an improved locust optimization algorithm (IGOA);

and optimizing the super-parameter learning rate and the number of hidden layer nodes in the LSTM algorithm by using an improved locust optimization algorithm (IGOA), and finding a group of parameters which minimize the loss function, thereby obtaining an optimal LSTM prediction model.

Where N 'is the number of samples in the test dataset, y' _i Is the actual value of the test data set,is the corresponding predicted output value.

S5, predicting the cold load of the airport terminal by using the optimized IGOA-LSTM model.

And selecting proper characteristics as input variables of an IGOA-LSTM prediction model according to the operation result of the random forest algorithm, and taking the cold load of the airport terminal as the output variable of the model. The model is trained by means of actual airport building data, and an airport terminal cold load prediction model with excellent performance is finally built through comparison and analysis with other widely applied algorithms, so that accurate airport terminal cold load prediction values can be obtained in actual application.

In still another embodiment of the present invention, a system for predicting a cold load of an airport terminal based on an RF-IGOA-LSTM model is provided, where the system can be used to implement the method for predicting a cold load of an airport terminal based on an RF-IGOA-LSTM model described above, and specifically, the system for predicting a cold load of an airport terminal based on an RF-IGOA-LSTM model includes a data module, an extraction module, an improvement module, an optimization module, and a prediction module.

The data module divides airport data according to seasons according to the characteristic that the cold load of the airport terminal changes along with weather, and data preprocessing is achieved;

The extraction module is used for extracting characteristics of the obtained data by using a random forest;

the improvement module is used for introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm;

the optimization module optimizes the super parameters in the LSTM algorithm by utilizing the improved locust optimization algorithm to obtain an IGOA-LSTM prediction model;

and the prediction module inputs the data into the optimized IGOA-LSTM model to realize the prediction of the cold load of the airport terminal.

In yet another embodiment of the present invention, a terminal device is provided, the terminal device including a processor and a memory, the memory for storing a computer program, the computer program including program instructions, the processor for executing the program instructions stored by the computer storage medium. The processor may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processor, digital signal processor (Digital Signal Processor, DSP), application specific integrated circuit (Application Specific Integrated Circuit, ASIC), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware components, etc., which are the computational core and control core of the terminal adapted to implement one or more instructions, in particular to load and execute one or more instructions to implement the corresponding method flow or corresponding functions; the processor according to the embodiment of the invention can be used for the operation of the cold load prediction method of the airport terminal based on the RF-IGOA-LSTM model, and comprises the following steps:

Dividing airport data according to seasons according to the characteristic that the cold load of airport terminal changes along with climate, and realizing data preprocessing; extracting features of the obtained data by using a random forest; introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm; optimizing the super parameters in the LSTM algorithm by using the improved locust optimization algorithm to obtain an IGOA-LSTM prediction model; and inputting the obtained data into an optimized IGOA-LSTM model to realize the prediction of the cold load of the airport terminal.

Referring to fig. 6, the terminal device is a computer device, and the computer device 60 of this embodiment includes: a processor 61, a memory 62, and a computer program 63 stored in the memory 62 and executable on the processor 61, the computer program 63 when executed by the processor 61 implements the reservoir inversion wellbore fluid composition calculation method of the embodiment, and is not described in detail herein to avoid repetition. Alternatively, the computer program 63 when executed by the processor 61 implements the functions of the various models/units in the RF-IGOA-LSTM model based airport terminal cold load prediction system, and is not described in detail herein to avoid repetition.

The computer device 60 may be a desktop computer, a notebook computer, a palm top computer, a cloud server, or the like. Computer device 60 may include, but is not limited to, a processor 61, a memory 62. It will be appreciated by those skilled in the art that fig. 6 is merely an example of a computer device 60 and is not intended to be limiting of the computer device 60, and may include more or fewer components than shown, or may combine certain components, or different components, e.g., a computer device may also include an input-output device, a network access device, a bus, etc.

The processor 61 may be a central processing unit (Central Processing Unit, CPU), but may also be other general purpose processors, digital signal processors (Digital Signal Processor, DSP), application specific integrated circuits (Application Specific Integrated Circuit, ASIC), field programmable gate arrays (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, or the like. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

The memory 62 may be an internal storage unit of the computer device 60, such as a hard disk or memory of the computer device 60. The memory 62 may also be an external storage device of the computer device 60, such as a plug-in hard disk provided on the computer device 60, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), or the like.

Further, the memory 62 may also include both internal storage units and external storage devices of the computer device 60. The memory 62 is used to store computer programs and other programs and data required by the computer device. The memory 62 may also be used to temporarily store data that has been output or is to be output.

Referring to fig. 7, the terminal device is a chip, and the chip 600 of this embodiment includes a processor 622, which may be one or more in number, and a memory 632 for storing a computer program executable by the processor 622. The computer program stored in memory 632 may include one or more modules each corresponding to a set of instructions. Further, the processor 622 may be configured to execute the computer program to perform the above-described method of cold load prediction for airport terminal based on the RF-IGOA-LSTM model.

In addition, chip 600 may further include a power supply component 626 and a communication component 650, where power supply component 626 may be configured to perform power management of chip 600, and communication component 650 may be configured to enable communication of chip 600, e.g., wired or wireless communication. In addition, the chip 600 may also include an input/output (I/O) interface 658. Chip 600 may operate based on an operating system stored in memory 632.

In a further embodiment of the present invention, the present invention also provides a storage medium, in particular, a computer readable storage medium (Memory), which is a Memory device in a terminal device, for storing programs and data. It will be appreciated that the computer readable storage medium herein may include both a built-in storage medium in the terminal device and an extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space storing an operating system of the terminal. Also stored in the memory space are one or more instructions, which may be one or more computer programs (including program code), adapted to be loaded and executed by the processor. The computer readable storage medium may be a high-speed RAM Memory or a Non-Volatile Memory (Non-Volatile Memory), such as at least one magnetic disk Memory.

One or more instructions stored in a computer-readable storage medium may be loaded and executed by a processor to implement the respective steps of the method for cold load prediction in the above-described embodiments with respect to an airport terminal based on an RF-IGOA-LSTM model; one or more instructions in a computer-readable storage medium are loaded by a processor and perform the steps of:

dividing airport data according to seasons according to the characteristic that the cold load of airport terminal changes along with climate, and realizing data preprocessing; extracting features of the obtained data by using a random forest; introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm; optimizing the super parameters in the LSTM algorithm by using the improved locust optimization algorithm to obtain an IGOA-LSTM prediction model; and inputting the obtained data into an optimized IGOA-LSTM model to realize the prediction of the cold load of the airport terminal.

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

According to the invention, experimental verification is carried out by collecting real cold load operation data of june to September of a airport terminal building, indoor and outdoor environments, personnel parameters and the like, and model training and testing are carried out, and meanwhile, comparison research on an optimization algorithm, a built prediction model and other algorithms widely applied is carried out, so that the method can accurately obtain the change condition of the cold load of a relevant airport building in practical application, is beneficial to realizing the online optimization control of an air conditioning system of the airport building, and provides very effective data support and theoretical reference for the energy-saving optimization management of the type of airport building and the formulation of an operation strategy.

According to the energy utilization characteristics of the existing data set, the number of the adopted training data samples is the historical data of each hour from early ten to evening ten in the month of June to September of the airport terminal, and the data is 1464 groups in total. The latter 190 sets of data were used as test samples and the remaining data were used as training data sets for the model. The final evaluation index is the Mean Absolute Error (MAE), mean Absolute Percent Error (MAPE), root Mean Square Error (RMSE) and determination coefficient (R-square) of the predicted value and the actual value, and the formula is as follows:

wherein,and y _i Respectively the actual and the predicted airport cooling load, < >>For average predicted cooling load, n is the total number of samples for a given test set and i is the number of samples for the test set.

The relevant evaluation indexes are shown in table 2, compared with the LSTM model modified by other modification algorithms and the LSTM model not modified.

Table 2 model prediction accuracy contrast

In summary, the method and the system for predicting the cold load of the airport terminal based on the RF-IGOA-LSTM model have excellent prediction precision and model generalization capability under a small sample, and the IGOA algorithm has better optimization capability than other French algorithm on the cold load prediction of the large airport terminal, can timely and accurately reflect the equipment operation and the energy consumption condition of the cold supply system of the large airport terminal in practical application, and provide effective data reference for the cold load prediction of the large airport terminal.

It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional units and modules is illustrated, and in practical application, the above-described functional distribution may be performed by different functional units and modules according to needs, i.e. the internal structure of the apparatus is divided into different functional units or modules to perform all or part of the above-described functions. The functional units and modules in the embodiment may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit, where the integrated units may be implemented in a form of hardware or a form of a software functional unit. In addition, specific names of the functional units and modules are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present application. The specific working process of the units and modules in the above system may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.

In the foregoing embodiments, the descriptions of the embodiments are emphasized, and in part, not described or illustrated in any particular embodiment, reference is made to the related descriptions of other embodiments.

Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus/terminal and method may be implemented in other manners. For example, the apparatus/terminal embodiments described above are merely illustrative, e.g., the division of the modules or units is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection via interfaces, devices or units, which may be in electrical, mechanical or other forms.

The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.

The integrated modules/units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the present invention may implement all or part of the flow of the method of the above embodiment, or may be implemented by a computer program to instruct related hardware, where the computer program may be stored in a computer readable storage medium, and when the computer program is executed by a processor, the computer program may implement the steps of each of the method embodiments described above. Wherein the computer program comprises computer program code which may be in source code form, object code form, executable file or some intermediate form etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a usb disk, a removable hard disk, a magnetic disk, an optical disk, a computer Memory, a Read-Only Memory (ROM), a Random-Access Memory (RAM), an electrical carrier wave signal, a telecommunications signal, a software distribution medium, etc., it should be noted that the content of the computer readable medium may be appropriately increased or decreased according to the requirements of legislation and patent practice in jurisdictions, such as in some jurisdictions, according to the legislation and patent practice, the computer readable medium does not include electrical carrier wave signals and telecommunications signals.

The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.

These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.

The above is only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited by this, and any modification made on the basis of the technical scheme according to the technical idea of the present invention falls within the protection scope of the claims of the present invention.

Claims (10)

1. The cold load prediction method of the airport terminal based on the RF-IGOA-LSTM model is characterized by comprising the following steps of:

s1, dividing airport data according to seasons according to the characteristic that the cold load of an airport terminal changes along with weather, and realizing data preprocessing;

s2, carrying out feature extraction on the data obtained in the step S1 by using a random forest;

s3, introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm;

s4, optimizing the super parameters in the LSTM algorithm by using the locust optimization algorithm improved in the step S3 to obtain an IGOA-LSTM prediction model;

S5, inputting the data obtained in the step S2 into the IGOA-LSTM model optimized in the step S4, and predicting the cold load of the airport terminal.

2. The method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model of claim 1, wherein the data is padded using linear interpolation during the data preprocessing and then normalized using max-min.

3. The method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model according to claim 1, wherein the data obtained in the step S1 is extracted by using random forests, and parameters with weight values larger than 0.8 are selected.

4. The method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model according to claim 3, wherein the feature extraction of data using random forests is specifically:

generating n training subsets by using a bootstrap sampling method according to an original training set, wherein each training set has m samples in total;

model training is carried out on the n training subsets, and n decision tree models are generated;

for the feature attribute of a single decision tree model, selecting the strongest segmentation capability feature from all features by using a minimum variance criterion to perform node segmentation;

Under the condition of no pruning, integrating each tree into a random forest after the maximum growth, and then calculating the influence of the data outside the bag on the model by using a mean square error;

voting on all decision trees, i.e. the importance of the input variables is represented by reduction of MSE, and the mean square residual sequence is obtained by using the out-of-bag data [ MSE1, MSE2, ··, MSEk ].

5. The method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model according to claim 1, wherein the position update formula of the locust individual introducing the cauchy-gaussian factor to improve the traditional locust optimization algorithm is as follows:

wherein,for the random position of the kth locust on the d dimension, gaussian is Gaussian operator, +.>For the optimal position of the t generation in the d-th dimension,>is a random position of the t generation in the d dimension.

6. The method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model according to claim 1, wherein the method is characterized in that the super parameter learning rate and the hidden layer node number in the LSTM algorithm are optimized by using an improved locust optimization algorithm, and a group of parameters which minimize a loss function are found to obtain the IGOA-LSTM prediction model, and the method is specifically as follows:

Where N 'is the number of samples in the test dataset, y' _i Is the actual value of the test data set,is the corresponding predicted output value.

7. The method for predicting the cold load of an airport terminal based on an RF-IGOA-LSTM model according to claim 1, wherein the predicting the cold load of the airport terminal specifically comprises:

according to the result of random forest algorithm operation, selecting the corresponding characteristic as an input variable of an IGOA-LSTM prediction model, and taking the cold load of an airport terminal as an output variable of the IGOA-LSTM prediction model; training an IGOA-LSTM prediction model by means of actual airport building data to obtain accurate airport terminal cold load prediction values.

8. An RF-IGOA-LSTM model based cold load prediction system for airport terminal, comprising:

the data module divides airport data according to seasons according to the characteristic that the cold load of the airport terminal changes along with weather, so as to realize data preprocessing;

the extraction module is used for extracting characteristics of the obtained data by using a random forest;

the improvement module is used for introducing a Cauchy-Gaussian factor to improve a traditional locust optimization algorithm;

the optimization module optimizes the super parameters in the LSTM algorithm by utilizing the improved locust optimization algorithm to obtain an IGOA-LSTM prediction model;

And the prediction module inputs the data into the optimized IGOA-LSTM model to realize the prediction of the cold load of the airport terminal.

9. A chip is characterized in that,

a memory having a computer program stored thereon;

a processor for executing the computer program in the memory to implement the steps of the method of any one of claims 1-7.

10. An electronic device, characterized in that,

comprising a chip as claimed in claim 9.