CN112906871A - Temperature prediction method and system based on hybrid multilayer neural network model - Google Patents

Temperature prediction method and system based on hybrid multilayer neural network model Download PDF

Info

Publication number
CN112906871A
CN112906871A CN202110315016.1A CN202110315016A CN112906871A CN 112906871 A CN112906871 A CN 112906871A CN 202110315016 A CN202110315016 A CN 202110315016A CN 112906871 A CN112906871 A CN 112906871A
Authority
CN
China
Prior art keywords
neural network
network model
layer
hybrid multi
temperature prediction
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202110315016.1A
Other languages
Chinese (zh)
Inventor
傅尊伟
张聪聪
马振明
王政
侯宪明
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Linyi University
Original Assignee
Linyi University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Linyi University filed Critical Linyi University
Priority to CN202110315016.1A priority Critical patent/CN112906871A/en
Publication of CN112906871A publication Critical patent/CN112906871A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/043—Architecture, e.g. interconnection topology based on fuzzy logic, fuzzy membership or fuzzy inference, e.g. adaptive neuro-fuzzy inference systems [ANFIS]
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q10/00—Administration; Management
    • G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Business, Economics & Management (AREA)
  • Health & Medical Sciences (AREA)
  • Strategic Management (AREA)
  • Human Resources & Organizations (AREA)
  • Economics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Mathematical Analysis (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computational Mathematics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Fuzzy Systems (AREA)
  • Development Economics (AREA)
  • Automation & Control Theory (AREA)
  • Game Theory and Decision Science (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Marketing (AREA)
  • Operations Research (AREA)
  • Quality & Reliability (AREA)
  • Tourism & Hospitality (AREA)

Abstract

The invention belongs to the field of artificial intelligence, and provides a temperature prediction method and system based on a hybrid multilayer neural network model. The temperature prediction method comprises the steps of obtaining weather information of a set historical time period of a set area; inputting the obtained weather information into a trained hybrid multilayer neural network model, and outputting a predicted value of the average temperature of a set area; the number of layers of the mixed multilayer neural network model at least comprises two layers, the first layer is formed by fuzzy neurons based on clustering, the continuous layers behind the first layer are all constructed by polynomial neurons, and the polynomial neurons of the continuous layers behind the first layer are obtained by screening through an evolutionary selection strategy.

Description

Temperature prediction method and system based on hybrid multilayer neural network model
Technical Field
The invention belongs to the field of artificial intelligence, and particularly relates to a temperature prediction method and system based on a hybrid multilayer neural network model.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
Predictive modeling and pattern classification becomes more challenging when encountering problems related to non-linearity, high dimensions, measurement inaccuracies and high order dynamics of data, and diversification of application scenarios. Hybrid modeling using the synergistic effect of multiple technologies is one of the effective methods to solve such problems. As an important component of computational intelligence research, neural networks have become an effective tool for studying hybrid architectures with adaptive learning capabilities. Neural networks are of many types, including classical neural networks (e.g., multilayer perceptrons), radial basis function neural networks, polynomial neural networks, and the like.
The original polynomial neural network only comprises neurons consisting of simple polynomials or linear functions, and the structure of the original polynomial neural network is simple and flexible but cannot well reflect and represent complex structures in data. Current polynomial neural network research focuses on replacing all nodes in a polynomial neural network with more complex neurons to enhance the nonlinear fitting capabilities of the model, but this also increases the structural complexity of the model. Particularly, as the number of network layers increases, the computational efficiency of the model is reduced more obviously, and meanwhile, the prediction accuracy of the model cannot be guaranteed. The performance of a polynomial neural network and its similar structural models depends on the combination of input variables, so the selection strategy of neurons in constructing each layer of the network has a very important influence on the prediction performance of the model.
The temperature prediction of the set area has important significance for arranging the work and life (such as agricultural production, military operation and the like) of the corresponding area. However, the inventor finds that the calculation efficiency of the temperature prediction of the set area is low and the temperature prediction accuracy cannot be guaranteed due to the complexity of the neural network at present.
Disclosure of Invention
In order to solve the technical problems in the background art, the invention provides a temperature prediction method and a temperature prediction system based on a hybrid multilayer neural network model, which have high calculation efficiency and strong generalization capability of the model, and can ensure the prediction accuracy of the average temperature prediction value of a set region.
In order to achieve the purpose, the invention adopts the following technical scheme:
the first aspect of the invention provides a temperature prediction method based on a hybrid multilayer neural network model.
A temperature prediction method based on a hybrid multilayer neural network model comprises the following steps:
acquiring weather information of a set historical time period of a set area;
inputting the obtained weather information into a trained hybrid multilayer neural network model, and outputting a predicted value of the average temperature of a set area;
the number of layers of the mixed multilayer neural network model at least comprises two layers, the first layer is formed by fuzzy neurons based on clustering, the continuous layers behind the first layer are all constructed by polynomial neurons, and the polynomial neurons of the continuous layers behind the first layer are obtained by screening through an evolutionary selection strategy.
Further, the weather information comprises a maximum temperature, a minimum temperature, a dew point, precipitation, sea level air pressure, standard air pressure, visibility, wind speed and a maximum wind speed.
Further, in the hybrid multi-layer neural network model, the input of the successive layers after the first layer is jointly determined by the output of the previous layer and the evolutionary selection strategy.
Further, in the first layer of the hybrid multilayer neural network model, a K-Means clustering algorithm is used for determining the center of a cluster, a Gaussian function is used as a membership function, the membership is calculated, the membership is standardized, and a linear function is used as a post part for expressing fuzzy rules.
Further, before training the hybrid multi-layer neural network model, the method further comprises:
acquiring weather information of a set historical time period of a set area and forming a data set;
the dataset was partitioned using five-fold cross validation.
Further, the evolutionary selection strategy is a championship selection strategy, the output of the previous layer is used as the candidate input of the current layer, the adaptive value of each current neuron is calculated, and then initial selection is carried out according to the adaptive value; the adaptive value is the sum of squares of differences between the measured average temperature and the predicted average temperature corresponding to all weather information training data.
Further, the evolutionary selection strategy is an optimal reservation selection or random competition selection strategy.
A second aspect of the invention provides a temperature prediction system based on a hybrid multi-layer neural network model.
A hybrid multi-layer neural network model-based temperature prediction system, comprising:
the weather information acquisition module is used for acquiring weather information of a set historical time period of a set region;
the average temperature prediction module is used for inputting the acquired weather information into the trained hybrid multilayer neural network model and outputting a predicted value of the average temperature of the set region;
the number of layers of the mixed multilayer neural network model at least comprises two layers, the first layer is formed by fuzzy neurons based on clustering, the continuous layers behind the first layer are all constructed by polynomial neurons, and the polynomial neurons of the continuous layers behind the first layer are obtained by screening through an evolutionary selection strategy.
A third aspect of the invention provides a computer-readable storage medium.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the hybrid multi-layer neural network model-based temperature prediction method as set forth above.
A fourth aspect of the invention provides a computer apparatus.
A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps in the hybrid multi-layer neural network model-based temperature prediction method as described above when executing the program.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a temperature prediction method based on a hybrid multilayer neural network model, which aims to solve the problems that the calculation efficiency of temperature prediction of a set area is low and the temperature prediction accuracy cannot be ensured.
Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate exemplary embodiments of the invention and together with the description serve to explain the invention and not to limit the invention.
FIG. 1 is a flow chart of a temperature prediction method based on a hybrid multi-layer neural network model according to an embodiment of the present invention;
FIG. 2 is a schematic structural diagram of a hybrid multi-layer neural network model provided by an embodiment of the present invention;
FIG. 3 is a tournament selection policy step provided by an embodiment of the invention;
fig. 4 is a schematic structural diagram of a temperature prediction system based on a hybrid multi-layer neural network model according to an embodiment of the present invention.
Detailed Description
The invention is further described with reference to the following figures and examples.
It is to be understood that the following detailed description is exemplary and is intended to provide further explanation of the invention as claimed. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of exemplary embodiments according to the invention. As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, and it should be understood that when the terms "comprises" and/or "comprising" are used in this specification, they specify the presence of stated features, steps, operations, devices, components, and/or combinations thereof, unless the context clearly indicates otherwise.
Example one
As shown in fig. 1, the present embodiment provides a temperature prediction method based on a hybrid multi-layer neural network model, which specifically includes the following steps:
s101: and acquiring weather information of a set historical time period of a set region.
It should be noted here that the weather information of the present embodiment includes a maximum temperature, a minimum temperature, a dew point, precipitation, sea level air pressure, standard air pressure, visibility, wind speed, and a maximum wind speed.
In other embodiments, other amounts of weather information including at least a maximum temperature and a minimum temperature may be used for the weather information.
S102: and inputting the acquired weather information into the trained hybrid multilayer neural network model, and outputting a predicted value of the average temperature of the set region.
The number of layers of the mixed multilayer neural network model at least comprises two layers, the first layer is formed by fuzzy neurons based on clustering, the continuous layers behind the first layer are all constructed by polynomial neurons, and the polynomial neurons of the continuous layers behind the first layer are obtained by screening through an evolutionary selection strategy, as shown in fig. 2.
In fig. 2, the first layer of the hybrid multi-layer neural network model is composed of cluster-based fuzzy neurons (CFNs). A CFN is essentially a set of fuzzy rules. The antecedent of the fuzzy rule is determined by a clustering algorithm, and is explained by a K-Means algorithm:
s1021: determining the center of the cluster by using a K-Means clustering algorithm, using a Gaussian function as a membership function, and calculating the membership, wherein the membership based on the Gaussian function can be expressed as:
Figure BDA0002990821770000061
where | | | - | represents the euclidean distance, viIs the center of the ith cluster, σiRepresenting the width of the gaussian function. x is the number ofkIndicating the kth piece of data. If the input of the preset neuron is 2, xkIs a piece of 2-dimensional data (containing two of the 9 features, e.g. visibility and maximum wind speed).
S1022: the membership degree obtained in S1021 is normalized. The normalized membership value is between 0 and 1.
S1023: using a linear function as the part of representing the fuzzy rule back-part, the output based on clustering neurons can be expressed as:
Figure BDA0002990821770000062
wherein c represents the number of clustering centers, the general value range is 2-10, and fi(xk) Representing the ith linear function.
In fig. 2, successive layers after the first layer are each constructed using polynomial neurons PN, using the form of a quadratic polynomial as follows:
Figure BDA0002990821770000063
x'kalthough also two-dimensional data, represents an enhanced feature (the output of the previous layer as the input of the next layer). t represents the index of the polynomial neuron on the current layer (second and further layers), and n represents the dimension of the data, here 2.
X 'herein'kRefer toThe output of the upper layer network is generated by combining the following formulas:
Figure BDA0002990821770000071
where G represents the number of features of the input data of the current layer, for example: in the first layer G equals 9, which represents the number of input variables per neuron, here set to 2. The result of the combination of G and G is the number of neurons that are composed in the current layer.
In the hybrid multi-layer neural network model, the input of the continuous layer after the first layer is jointly determined by the output of the previous layer and the evolutionary selection strategy.
It is noted that the training and testing input data for each layer are different, but their respective outputs are the same as the original training and testing output data.
As shown in FIG. 3, the evolutionary selection strategy is a tournament selection strategy. The neuron selection procedure based on the tournament selection strategy is as follows:
step (1), the output of the previous layer is used as the candidate input of the current layer, the adaptive value of each current neuron is calculated, and then initial selection is carried out according to the adaptive value; the adaptive value is the sum of squares of differences between the measured average temperature and the predicted average temperature corresponding to all weather information training data.
Figure BDA0002990821770000072
Where N is the number of training data, for example 1287. y isiWhich represents the average temperature measured and is,
Figure BDA0002990821770000073
is the predicted average temperature.
After the initial selection in step (2), randomly selecting some individuals from the neuron set to construct the tournament pool and enable the individuals to participate in the tournament, wherein the number of the selected neurons is the same as the size of the tournament pool.
And (3) selecting the neurons from the tournament pool according to the neuron selection probability and the selection threshold. The selection probability of the neuron is:
SPi＝p(1-p)i,i＝0,1,…,TS-1
wherein p is a neuron selection parameter between 0 and 1, and is set to 0.5, and TS is the scale of the championship game (namely, several individuals participate in the game). SPiThe selection probability for the ith neuron. The selection threshold is generally a random number of 0 to 1.
Step (4) selects the best several nodes in the race, usually only the best one.
Repeating the above steps (1) -4 until the number of selected neurons satisfies the predefined maximum number of nodes in each layer, here set to 20.
In other embodiments, the evolutionary selection strategy may also employ a best-retention selection or random competition selection strategy for screening neurons.
In a specific implementation, after each layer of network is built, the termination condition is checked, and if the network reaches a predetermined number of layers (here, 5 layers), the hybrid multi-layer neural network model is built.
In a specific implementation, before training the hybrid multi-layer neural network model, the method further includes:
acquiring weather information of a set historical time period of a set area and forming a data set;
the dataset was partitioned using five-fold cross validation.
For example: given a data set of weather information collected in a certain area. The data set contains 1609 pieces of data in total. Each piece of data contains 9 features, which are respectively: maximum temperature, minimum temperature, dew point, precipitation, sea level barometric pressure, standard barometric pressure, visibility, wind speed, maximum wind speed. The data was partitioned using five-fold cross-validation, i.e., the weather data set was partitioned into five pieces, four of which served as training (approximately 1287 pieces of data). In the first layer construction of the network, 1287 pieces of data are directly used as training input of the first layer, and since each piece of data has 9 features, the actual data is a 9 x 1287 matrix.
The temperature prediction method based on the hybrid multilayer neural network model of the embodiment is compared with the experimental results of other models on the same weather data set, as shown in table 1:
TABLE 1 comparison of the results
Model (model) Multilayer perceptron Support vector regression Linear regression The invention
Prediction error 1.4974 1.5762 1.5702 1.3109
As can be seen from table 1, in the present embodiment, the advantages of the fuzzy neuron and the polynomial neuron based on clustering are combined, and the effective neuron selection strategy based on evolutionary selection is utilized, so that the prediction performance of the model is significantly improved, that is, the prediction error is reduced.
Example two
As shown in fig. 4, the embodiment provides a temperature prediction system based on a hybrid multi-layer neural network model, which specifically includes the following modules:
a weather information acquisition module 11, configured to acquire weather information of a set historical time period of a set area;
the average temperature prediction module 12 is used for inputting the acquired weather information into the trained hybrid multilayer neural network model and outputting a predicted value of the average temperature of the set region;
the number of layers of the mixed multilayer neural network model at least comprises two layers, the first layer is formed by fuzzy neurons based on clustering, the continuous layers behind the first layer are all constructed by polynomial neurons, and the polynomial neurons of the continuous layers behind the first layer are obtained by screening through an evolutionary selection strategy.
It should be noted that, each module in the temperature prediction system based on the hybrid multi-layer neural network model of the present embodiment corresponds to each step in the temperature prediction method based on the hybrid multi-layer neural network model of the first embodiment one by one, and the specific implementation process is the same, and will not be described again here.
EXAMPLE III
The present embodiment provides a computer-readable storage medium, on which a computer program is stored, which when executed by a processor implements the steps in the hybrid multi-layer neural network model-based temperature prediction method as described above.
Example four
The present embodiment provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the hybrid multi-layer neural network model-based temperature prediction method as described above.
As will be appreciated by one skilled in the art, embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of a hardware embodiment, a software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams 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.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention, and various modifications and changes may be made by those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (10)

1.一种基于混合多层神经网络模型的温度预测方法，其特征在于，包括：1. a temperature prediction method based on a hybrid multi-layer neural network model, is characterized in that, comprising: 获取设定地区的设定历史时间段的天气信息；Obtain weather information for a set historical time period in a set area; 将获取的天气信息输入至训练完成的混合多层神经网络模型中，输出设定地区平均温度预测值；Input the obtained weather information into the trained hybrid multi-layer neural network model, and output the predicted value of the average temperature in the set area; 其中，混合多层神经网络模型的层数至少包括两层，第一层基于聚类的模糊神经元构成，第一层之后的连续层均使用多项式神经元构造，第一层之后的连续层的多项式神经元使用进化选择策略筛选得到。Among them, the number of layers of the hybrid multi-layer neural network model includes at least two layers. The first layer is composed of fuzzy neurons based on clustering, and the continuous layers after the first layer are constructed using polynomial neurons. Polynomial neurons were selected using an evolutionary selection strategy. 2.如权利要求1所述的基于混合多层神经网络模型的温度预测方法，其特征在于，所述天气信息包括最高温度、最低温度、结露点、降水量、海平面气压、标准气压、能见度、风速及最大风速。2. The temperature prediction method based on a hybrid multi-layer neural network model according to claim 1, wherein the weather information includes maximum temperature, minimum temperature, dew point, precipitation, sea level air pressure, standard air pressure, visibility , wind speed and maximum wind speed. 3.如权利要求1所述的基于混合多层神经网络模型的温度预测方法，其特征在于，在所述混合多层神经网络模型中，第一层之后的连续层的输入，由上一层的输出和进化选择策略共同决定。3. The temperature prediction method based on a hybrid multi-layer neural network model according to claim 1, wherein, in the hybrid multi-layer neural network model, the input of successive layers after the first layer is determined by the upper layer. The output of , and the evolutionary selection strategy are jointly determined. 4.如权利要求1所述的基于混合多层神经网络模型的温度预测方法，其特征在于，在所述混合多层神经网络模型的第一层中，使用K-Means聚类算法确定聚类的中心并利用高斯函数作为隶属度函数并计算隶属度，对隶属度进行标准化处理，使用线性函数作表示模糊规则后件部分。4. The temperature prediction method based on the hybrid multi-layer neural network model as claimed in claim 1, wherein, in the first layer of the hybrid multi-layer neural network model, K-Means clustering algorithm is used to determine clusters and use Gaussian function as the membership function and calculate the membership, standardize the membership, and use the linear function as the posterior part of the fuzzy rule. 5.如权利要求1所述的基于混合多层神经网络模型的温度预测方法，其特征在于，在训练混合多层神经网络模型之前，还包括：5. the temperature prediction method based on the hybrid multi-layer neural network model as claimed in claim 1, is characterized in that, before training the hybrid multi-layer neural network model, also comprises: 获取设定地区的设定历史时间段的天气信息并形成数据集；Obtain the weather information of the set historical time period in the set area and form a data set; 使用五折交叉验证对数据集进行划分。The dataset was partitioned using five-fold cross-validation. 6.如权利要求1所述的基于混合多层神经网络模型的温度预测方法，其特征在于，所述进化选择策略为锦标赛选择策略，由上一层的输出作为当前层的候选输入，计算当前每个神经元的适应值，然后根据适应值进行初始选择；其中，适应值为所有天气信息训练数据对应的测得平均温度与预测的平均温度的之差的平方累加和。6. The temperature prediction method based on a hybrid multi-layer neural network model as claimed in claim 1, wherein the evolutionary selection strategy is a championship selection strategy, and the output of the previous layer is used as the candidate input of the current layer, and the current layer is calculated. The fitness value of each neuron is then initially selected according to the fitness value; wherein, the fitness value is the cumulative sum of the squares of the difference between the measured average temperature and the predicted average temperature corresponding to all the weather information training data. 7.如权利要求1所述的基于混合多层神经网络模型的温度预测方法，其特征在于，所述进化选择策略为最佳保留选择或随机竞争选择策略。7 . The temperature prediction method based on a hybrid multi-layer neural network model according to claim 1 , wherein the evolutionary selection strategy is an optimal reservation selection or a random competition selection strategy. 8 . 8.一种基于混合多层神经网络模型的温度预测系统，其特征在于，包括：8. A temperature prediction system based on a hybrid multi-layer neural network model, comprising: 天气信息获取模块，其用于获取设定地区的设定历史时间段的天气信息；A weather information acquisition module, which is used to acquire weather information of a set historical time period in a set area; 平均温度预测模块，其用于将获取的天气信息输入至训练完成的混合多层神经网络模型中，输出设定地区平均温度预测值；an average temperature prediction module, which is used to input the obtained weather information into the trained hybrid multi-layer neural network model, and output the average temperature prediction value of the set area; 其中，混合多层神经网络模型的层数至少包括两层，第一层基于聚类的模糊神经元构成，第一层之后的连续层均使用多项式神经元构造，第一层之后的连续层的多项式神经元使用进化选择策略筛选得到。Among them, the number of layers of the hybrid multi-layer neural network model includes at least two layers. The first layer is composed of fuzzy neurons based on clustering, and the continuous layers after the first layer are constructed using polynomial neurons. Polynomial neurons were selected using an evolutionary selection strategy. 9.一种计算机可读存储介质，其上存储有计算机程序，其特征在于，该程序被处理器执行时实现如权利要求1-7中任一项所述的基于混合多层神经网络模型的温度预测方法中的步骤。9. A computer-readable storage medium on which a computer program is stored, characterized in that, when the program is executed by a processor, the hybrid multi-layer neural network model based on the Steps in a temperature prediction method. 10.一种计算机设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述处理器执行所述程序时实现如权利要求1-7中任一项所述的基于混合多层神经网络模型的温度预测方法中的步骤。10. A computer device, comprising a memory, a processor and a computer program stored in the memory and running on the processor, wherein the processor implements any of claims 1-7 when the processor executes the program. Steps in a temperature prediction method based on a hybrid multi-layer neural network model.
CN202110315016.1A 2021-03-24 2021-03-24 Temperature prediction method and system based on hybrid multilayer neural network model Pending CN112906871A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110315016.1A CN112906871A (en) 2021-03-24 2021-03-24 Temperature prediction method and system based on hybrid multilayer neural network model

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110315016.1A CN112906871A (en) 2021-03-24 2021-03-24 Temperature prediction method and system based on hybrid multilayer neural network model

Publications (1)

Publication Number Publication Date
CN112906871A true CN112906871A (en) 2021-06-04

Family

ID=76106252

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110315016.1A Pending CN112906871A (en) 2021-03-24 2021-03-24 Temperature prediction method and system based on hybrid multilayer neural network model

Country Status (1)

Country Link
CN (1) CN112906871A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114861775A (en) * 2022-04-21 2022-08-05 南京信息工程大学 A hybrid neural network model weather classification method after feature selection and cluster analysis
CN116595806A (en) * 2023-07-14 2023-08-15 江西师范大学 Self-adaptive temperature data complement method

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102622515A (en) * 2012-02-21 2012-08-01 北京联合大学 Weather prediction method
CN103426026A (en) * 2013-09-10 2013-12-04 信阳师范学院 Method for forecasting hybrid neural network and recognizing scenic spot meteorological elements
CN110991690A (en) * 2019-10-17 2020-04-10 宁波大学 Multi-time wind speed prediction method based on deep convolutional neural network
AU2020103741A4 (en) * 2020-11-27 2021-02-11 J. Arunnehru Feed forward neural networks combined with extreme learning machine approach for large weather data

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102622515A (en) * 2012-02-21 2012-08-01 北京联合大学 Weather prediction method
CN103426026A (en) * 2013-09-10 2013-12-04 信阳师范学院 Method for forecasting hybrid neural network and recognizing scenic spot meteorological elements
CN110991690A (en) * 2019-10-17 2020-04-10 宁波大学 Multi-time wind speed prediction method based on deep convolutional neural network
AU2020103741A4 (en) * 2020-11-27 2021-02-11 J. Arunnehru Feed forward neural networks combined with extreme learning machine approach for large weather data

Non-Patent Citations (8)

* Cited by examiner, † Cited by third party
Title
C.ZHANG ET AL.: "《Design of Reinforced Hybrid Fuzzy Rule-Based》", 《IEEE TRANSACTION ON FUZZY SYSTEM》 *
C.ZHANG ET AL.: "《Design of Reinforced Hybrid Fuzzy Rule-Based》", 《IEEE TRANSACTION ON FUZZY SYSTEM》, 20 August 2020 (2020-08-20), pages 3293 - 3307 *
C.ZHANG ET AL.: "Design of Reinforced Hybrid Fuzzy Rule-BasedNeural Networks Driven to Inhomogeneous Neurons and Tournament Selection", 《IEEE TRANSACTIONS ON FUZZY SYSTEMS》, vol. 29, no. 11, pages 3293 - 3307, XP011886044, DOI: 10.1109/TFUZZ.2020.3018190 *
C.ZHANG ET AL.: "Self-organized hybrid fuzzy neural networks driven with the aid of probability-based node selection and enhanced input strategy", 《NEUROCOMPUTING》, pages 471 - 489 *
ZHANG C ET AL.: "《Design of stabilized polynomial-based ensemble fuzzy neural networks based on heterogeneous neurons and synergy of multiple techniques》", 《INFORMATION SCIENCES》 *
ZHANG C ET AL.: "《Design of stabilized polynomial-based ensemble fuzzy neural networks based on heterogeneous neurons and synergy of multiple techniques》", 《INFORMATION SCIENCES》, 21 July 2020 (2020-07-21), pages 425 - 452 *
ZHANG C ET AL.: "《Self-organized hybrid fuzzy neural networks driven with the aid of probability-based node selection and enhanced input strategy》", 《NEUROCOMPUTING》 *
ZHANG C ET AL.: "《Self-organized hybrid fuzzy neural networks driven with the aid of probability-based node selection and enhanced input strategy》", 《NEUROCOMPUTING》, 12 September 2020 (2020-09-12), pages 471 - 489 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114861775A (en) * 2022-04-21 2022-08-05 南京信息工程大学 A hybrid neural network model weather classification method after feature selection and cluster analysis
CN114861775B (en) * 2022-04-21 2025-06-10 南京信息工程大学 Mixed neural network model weather classification method after feature selection and cluster analysis
CN116595806A (en) * 2023-07-14 2023-08-15 江西师范大学 Self-adaptive temperature data complement method
CN116595806B (en) * 2023-07-14 2023-10-10 江西师范大学 Self-adaptive temperature data complement method

Similar Documents

Publication Publication Date Title
WO2012177913A1 (en) Method and apparatus for a local competitive learning rule that leads to sparse connectivity
Ahmadi et al. Learning fuzzy cognitive maps using imperialist competitive algorithm
CN103226566A (en) System and method for optimizing teams
CN107223260B (en) Method for dynamically updating classifier complexity
CN108594793A (en) A kind of improved RBF flight control systems fault diagnosis network training method
Dhahri et al. Hierarchical multi-dimensional differential evolution for the design of beta basis function neural network
KR101901307B1 (en) Method, apparatus and computer-readable recording medium for classifying classes using deep neural network with weighted fuzzy membership function
CN112906871A (en) Temperature prediction method and system based on hybrid multilayer neural network model
CN114091597A (en) Countermeasure training method, device and equipment based on adaptive group sample disturbance constraint
CN117273105A (en) A module construction method and device for neural network models
Gupta et al. Optimizing deep neural network architecture: A Tabu search based approach
JPH11510936A (en) Optimization method of fuzzy rules by computer
CN118335189A (en) Single-cell deep clustering method fused with variational graph attention autoencoder
Worasucheep Forecasting currency exchange rates with an Artificial Bee Colony-optimized neural network
CN110109005B (en) Analog circuit fault testing method based on sequential testing
Kuo et al. Generalized part family formation through fuzzy self-organizing feature map neural network
CN116541767A (en) A method and system for predicting multivariate greenhouse environmental parameters based on graph neural network
Zhang et al. Intelligent Information Processing with Matlab
CN104915714A (en) Predication method and device based on echo state network (ESN)
CN114880538A (en) Attribute graph community detection method based on self-supervision
Liu et al. A new ART-counterpropagation neural network for solving a forecasting problem
CN109858520B (en) Multi-layer semi-supervised classification method
Yamada et al. Weight Features for Predicting Future Model Performance of Deep Neural Networks.
CN111666991A (en) Convolutional neural network-based pattern recognition method and device and computer equipment
CN103198357A (en) Optimized and improved fuzzy classification model construction method based on nondominated sorting genetic algorithm II (NSGA- II)

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination