CN112990591A

Movatterモバイル変換

Info

Publication number: CN112990591A
Application number: CN202110325811.9A
Authority: CN
Inventors: 邓必涛; 邓靖川; 冯海云; 杜晓丹; 李剑
Original assignee: Jiangxi Energy Big Data Co ltd
Current assignee: Jiangxi Energy Big Data Co ltd
Priority date: 2021-03-26
Filing date: 2021-03-26
Publication date: 2021-06-18

Abstract

Description

Multi-dimensional energy consumption data analysis method based on convolutional neural network and enterprise energy consumption prediction model

Technical Field

The invention belongs to the technical field of network analysis and prediction, and particularly relates to a multi-dimensional energy consumption data analysis method and an enterprise energy consumption prediction model based on a convolutional neural network.

Background

Convolutional Neural Networks (CNN) are a class of feed forward Neural Networks (fed forward Neural Networks) that contain convolution computations and have a deep structure, and are one of the representative algorithms for deep learning (deep learning). Convolutional Neural Networks have a feature learning (rendering) capability, and can perform Shift-Invariant classification (Shift-Invariant classification) on input information according to a hierarchical structure thereof, and are therefore also called "Shift-Invariant Artificial Neural Networks (SIANN)".

The study of convolutional neural networks began in the 80 to 90 s of the twentieth century, with time delay networks and LeNet-5 being the earliest convolutional neural networks that emerged; after the twenty-first century, with the introduction of deep learning theory and the improvement of numerical computing equipment, convolutional neural networks have been rapidly developed and applied to the fields of computer vision, natural language processing, and the like.

The convolutional neural network is constructed by imitating a visual perception (visual perception) mechanism of a living being, can perform supervised learning and unsupervised learning, and has the advantages that the convolutional neural network can learn grid-like topologic features such as pixels and audio with small calculation amount, has stable effect and has no additional feature engineering (feature engineering) requirement on data due to the fact that convolutional kernel parameter sharing in an implicit layer and sparsity of connection between layers.

In view of the defects of the traditional prediction method, a genetic algorithm is used for optimizing a wavelet neural network, and a wavelet neural network model based on the genetic algorithm is established for predicting the energy consumption of the Anshan iron and steel company.

Disclosure of Invention

The invention aims to provide a multi-dimensional energy consumption data analysis method and an enterprise energy consumption prediction model based on a convolutional neural network.

In order to achieve the purpose, the invention provides the following technical scheme: a multi-dimensional energy consumption data analysis method and an enterprise energy consumption prediction model based on a convolutional neural network comprise the following steps:

s1, constructing an enterprise energy consumption data set;

s2, constructing a convolutional neural network based on the enterprise energy consumption data;

s3, constructing an enterprise multi-azimuth energy consumption influence factor data set;

s4, training the constructed convolutional neural network by utilizing the enterprise multi-azimuth energy consumption influence factor data set, and adjusting parameters to evaluate to obtain an optimal convolutional neural network model of enterprise energy consumption;

s5, establishing an energy consumption prediction model based on the convolutional neural network, researching the influence of parameters on an energy consumption prediction structure, comparing the convolutional neural network with the optimal convolutional neural network model, and researching multi-azimuth energy consumption influence factors of enterprises.

Preferably, the enterprise energy consumption data in step S1 includes enterprise infrastructure energy consumption, enterprise production and processing energy consumption, and energy surplus generated and collected by enterprise energy recycling.

Preferably, in step S2, the convolutional neural network includes a neural network tree set inside the enterprise energy consumption data.

Preferably, the enterprise multi-azimuth energy consumption influencing factors in the step S3 include equipment service life, enterprise processing raw material quality, enterprise equipment maintenance aging information, and enterprise personnel and equipment operation methods.

Preferably, in the training in step S4, the energy consumption influencing factor is added to the neural network for training, and the variable factor in the energy consumption influencing factor is changed for gradual training.

Preferably, in the step S5, when energy consumption is predicted, different influencing factors may be input to perform energy consumption evaluation.

The invention has the beneficial effects that: according to the multi-dimensional energy consumption data analysis method based on the convolutional neural network and the enterprise energy consumption prediction model, the enterprise multi-azimuth multi-dimensional influence energy consumption factors are firstly trained through the convolutional neural network, energy consumption prediction is carried out through data input, and the energy consumption prediction efficiency is high and accurate.

Drawings

Fig. 1 is a schematic diagram of a multidimensional energy consumption data analysis method and an enterprise energy consumption prediction model based on a convolutional neural network provided by the invention.

Detailed Description

In order to further understand the contents, features and effects of the present invention, the following embodiments are illustrated and described in detail with reference to the accompanying drawings.

Referring to fig. 1, a detailed description will be given below of a convolutional neural network-based multidimensional energy consumption data analysis method and an enterprise energy consumption prediction model according to an embodiment of the present invention with reference to the drawings.

The multi-dimensional energy consumption data analysis method and the enterprise energy consumption prediction model based on the convolutional neural network comprise the following steps of:

s1, constructing an enterprise energy consumption data set;

Specifically, the enterprise energy consumption data in step S1 includes enterprise infrastructure energy consumption, enterprise production and processing energy consumption, and energy surplus generated and collected by enterprise energy recycling.

Specifically, in step S2, the convolutional neural network sets a neural network tree in the enterprise energy consumption data.

Specifically, the enterprise multi-azimuth energy consumption influencing factors in the step S3 include equipment service life, enterprise processing raw material quality, enterprise equipment maintenance aging information, and enterprise personnel and equipment operation methods.

Specifically, in the training in step S4, the energy consumption influencing factors are added to the neural network for training, and meanwhile, the variable factors in the energy consumption influencing factors are changed for gradual training.

Specifically, in the step S5, when energy consumption is predicted, different influencing factors may be input to perform energy consumption evaluation.

In the invention, the forecasting model is trained by using the foreseeable influence factors, and when energy consumption is forecasted, various data of an enterprise are input through the convolutional neural network, so that enterprise energy consumption forecasting data can be provided.

Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.

Claims

1. A multi-dimensional energy consumption data analysis method and an enterprise energy consumption prediction model based on a convolutional neural network are characterized by comprising the following steps:

s1, constructing an enterprise energy consumption data set;

2. The convolutional neural network-based multidimensional energy consumption data analysis method and the enterprise energy consumption prediction model as claimed in claim 1, wherein the enterprise energy consumption data in the step S1 includes enterprise infrastructure energy consumption, enterprise production and processing energy consumption, and energy consumption residue generated and collected by enterprise energy recycling.

3. The convolutional neural network based multidimensional energy consumption data analysis method and the enterprise energy consumption prediction model as claimed in claim 1, wherein the convolutional neural network in step S2 includes internal separately configured neural network trees in the enterprise energy consumption data.

4. The convolutional neural network-based multidimensional energy consumption data analysis method and the enterprise energy consumption prediction model as claimed in claim 1, wherein the enterprise multi-azimuth energy consumption influencing factors in the step S3 include equipment service life, quality of enterprise processing raw materials, enterprise equipment maintenance aging information, and enterprise personnel and equipment operation methods.

5. The convolutional neural network-based multidimensional energy consumption data analysis method and the enterprise energy consumption prediction model as claimed in claim 1, wherein in the training in step S4, the energy consumption influencing factors are added to the neural network for training, and meanwhile, variable factors in the energy consumption influencing factors are changed for gradual training.

6. The convolutional neural network-based multidimensional energy consumption data analysis method and the enterprise energy consumption prediction model as claimed in claim 1, wherein in the step S5, different influencing factors can be input for energy consumption assessment during energy consumption prediction.