CN113837420A - Power consumption prediction method, power consumption prediction system, and computer-readable storage medium - Google Patents

Power consumption prediction method, power consumption prediction system, and computer-readable storage medium Download PDF

Info

Publication number
CN113837420A
CN113837420A CN202010580966.2A CN202010580966A CN113837420A CN 113837420 A CN113837420 A CN 113837420A CN 202010580966 A CN202010580966 A CN 202010580966A CN 113837420 A CN113837420 A CN 113837420A
Authority
CN
China
Prior art keywords
power consumption
data
actual
unit time
per unit
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
CN202010580966.2A
Other languages
Chinese (zh)
Inventor
肖剑
杨菲
关瑞秋
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.)
Mitsubishi Electric China Co Ltd
Original Assignee
Mitsubishi Electric China Co Ltd
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 Mitsubishi Electric China Co Ltd filed Critical Mitsubishi Electric China Co Ltd
Priority to CN202010580966.2A priority Critical patent/CN113837420A/en
Publication of CN113837420A publication Critical patent/CN113837420A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q10/00Administration; Management
    • G06Q10/04Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION 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
    • G06Q50/00Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
    • G06Q50/06Energy or water supply
    • YGENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02PCLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
    • Y02P80/00Climate change mitigation technologies for sector-wide applications
    • Y02P80/10Efficient use of energy, e.g. using compressed air or pressurized fluid as energy carrier

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Economics (AREA)
  • Human Resources & Organizations (AREA)
  • Strategic Management (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Marketing (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Tourism & Hospitality (AREA)
  • Quality & Reliability (AREA)
  • Game Theory and Decision Science (AREA)
  • Operations Research (AREA)
  • Development Economics (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Public Health (AREA)
  • Water Supply & Treatment (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention provides a power consumption prediction method, which comprises the following steps: the prediction object produces or works according to the production plan or the work plan; acquiring actual production amount data or actual workload data of the prediction target in a predetermined period, and measuring actual power consumption value data per unit time, which is actual power consumption value data per unit time, of the prediction target in the predetermined period corresponding to the actual production amount data or actual workload data in the predetermined period; learning the actual production data or the actual workload data and the actual power consumption value data per unit time, and establishing a data model; when the prediction target is to perform production or work according to a new production plan or work plan, predicted power consumption value data per unit time, which is predicted power consumption value data per unit time corresponding to the new production plan or work plan, is predicted based on the data model.

Description

Power consumption prediction method, power consumption prediction system, and computer-readable storage medium
Technical Field
The present invention relates to a method for predicting power consumption of an electricity utility such as an enterprise plant or an electricity utility such as an air conditioner, a power consumption prediction system using the same, and a computer-readable storage medium.
Background
The current industrial or commercial electricity uses peak-valley time-of-use electricity price, which is basically divided into peak time period electricity consumption, ordinary time period electricity consumption and valley time period electricity consumption according to the price from high to low. Among them, the peak time period (7: 30-11: 30, 17: 00-21: 00) is very expensive, and the peak load shifting is adopted or other power utilization methods such as solar cells, photovoltaic cells, storage batteries and the like are adopted to replace the peak time period power utilization, so as to reduce the power utilization cost of the peak time period. Therefore, when a production plan is made and executed in an enterprise factory, power consumption corresponding to the production plan needs to be considered and managed to avoid or reduce peak time period power consumption, so that production cost is controlled, and reasonable power consumption is realized.
For example, patent document 1 discloses a method for monitoring and controlling energy costs of generation of a production lot, which measures energy consumption of each individual production step in a production process during execution of the production process and correlates the resulting energy consumption data with production specific data of a specific production lot, and adjusts a scheduling order of at least one production step by comparing an energy consumption profile with an energy cost profile to avoid consuming peak period electricity and/or consuming energy at low cost intervals.
Patent document 2 discloses a method for predicting an upper power limit (minimum power consumption) of an energy storage device, which determines an upper optimum power limit for a historical time period using historical data, and predicts an upper optimum power limit for a future use time period via a training neural network in combination with weather and building data.
Documents of the prior art
Patent document
Patent document 1: CN110032149A
Patent document 2: CN109670627A
Disclosure of Invention
Technical problem to be solved by the invention
In patent document 1, for considering the energy consumption of a production facility/line with respect to a specific product lot as a direct component of production, production data and energy consumption data and algorithms are associated and compared with energy costs, energy efficiency is improved by reducing the demand and fluctuation of electricity usage during peak periods, and energy costs are reduced by rearranging and optimizing production steps according to the energy costs. However, in this method, the energy consumption of a specific batch is measured during the execution of the production process of the batch, and the production data and the energy consumption data are associated by establishing a data model to optimize the production plan, so that there is a problem that the energy consumption corresponding to the production plan cannot be predicted before the production plan is formally executed. If the enterprise factory is able to predict the corresponding energy consumption before executing the production plan, peak electricity usage can be handled in advance, or an alternative manner of electricity usage can be taken to reduce costs, or the production plan can be adjusted to achieve cost reduction and green production.
In patent document 2, the future optimum power upper limit can be predicted from the historical data, but meter data (power meter data), weather data, building data, and the like are used, and the energy consumption according to the production plan cannot be predicted before the production plan is formally executed.
The present invention has been made in view of such circumstances, and an object thereof is to provide a power consumption prediction method, a power consumption prediction system, and a computer-readable storage medium, which can predict power consumption data of a current production plan from historical production data of a prediction target and corresponding power consumption data, and adjust or control the production plan according to the prediction result, thereby reducing peak-hour power consumption, reducing power consumption cost, and achieving optimal allocation of power resources.
Technical means for solving the technical problems
According to an aspect of an embodiment of the present invention, there is provided a power consumption prediction method including the steps of:
(1) the prediction object produces or works according to the production plan or the work plan;
(2) acquiring actual production amount data or actual workload data of the prediction target in a predetermined period, and measuring actual power consumption value data per unit time, which is actual power consumption value data per unit time, of the prediction target in the predetermined period corresponding to the actual production amount data or actual workload data in the predetermined period;
(3) learning the actual production data or the actual workload data and the actual power consumption value data per unit time, and establishing a data model;
(4) when the prediction target is to perform production or work according to a new production plan or work plan, predicted power consumption value data per unit time, which is predicted power consumption value data per unit time corresponding to the new production plan or work plan, is predicted based on the data model.
Alternatively, the predicted power consumption value data per unit time is supplied to a power management system to manage and control the actual power consumption of the prediction object.
Optionally, the measured actual production data or actual workload data and the actual power consumption value data per unit time are screened, and feature data is selected for establishing the data model.
Optionally, the feature data is analyzed to generate a plurality of analysis models including a linear regression model, a non-linear regression model, and an artificial neural network model.
Alternatively, the measured actual production amount data or actual workload data and the actual power consumption value data per unit time are substituted into an analysis model and compared with the predicted power consumption value data per unit time, and the plurality of analysis models are verified and evaluated, respectively, from which an analysis model for building the data model is selected.
Alternatively, the actual power consumption value data per unit time and the predicted power consumption value data per unit time indicate a peak value of power consumption per unit time of the prediction target.
Optionally, the predetermined period is one day, and the unit time is one minute.
Alternatively, the prediction object is a power consumption place including a factory, a building, a hospital, a market, and the like.
Alternatively, the prediction target is an electric power consumption device including an air conditioner and a motor.
According to another aspect of the embodiments of the present invention, there is also provided a power consumption prediction system for the above power consumption prediction method, the power consumption prediction system including:
the acquisition module is used for acquiring actual production volume data or actual workload data within a preset period when the prediction object carries out production or work according to a production plan or a work plan;
a measurement module configured to measure actual power consumption value data per unit time, which is actual power consumption value data per unit time, of the prediction target in the predetermined period, corresponding to the actual production amount data or the actual workload data in the predetermined period;
the data modeling module is used for learning the actual production data or the actual workload data and the actual power consumption value data per unit time and establishing a data model;
and a prediction module that predicts predicted power consumption value data per unit time, which is predicted power consumption value data per unit time corresponding to a new production plan or work plan, based on the data model when the prediction target is to produce or work according to the new production plan or work plan.
Optionally, the data modeling module comprises:
a feature data extraction unit that screens the measured actual production volume data or actual workload data and the actual power consumption value data per unit time, and selects feature data for creating the data model; and
an AI learning unit that learns the feature data, generates a plurality of analysis models including a linear regression model, a nonlinear regression model, and an artificial neural network model, and substitutes the measured actual production amount data or actual workload data and the actual power consumption value data per unit time into the analysis models to compare the actual production amount data or actual workload data with the predicted power consumption value data per unit time, thereby verifying and evaluating each of the plurality of analysis models, and selects an analysis model for creating the data model from the plurality of analysis models.
According to still another aspect of embodiments of the present invention, there is also provided a computer-readable storage medium storing a computer program for causing a computer to execute the above power consumption prediction method.
Effects of the invention
According to the power consumption prediction method of the present invention, the actual production amount data or the actual workload data of the prediction target in the predetermined period is used, the actual power consumption value data per unit time of the prediction target corresponding to the actual production amount data or the actual workload data in the predetermined period is measured, a data model associated with the two is established by learning, a new production plan or work plan is input, and the predicted power consumption value data per unit time corresponding to the new production plan or work plan is predicted based on the data model, so that the production plan can be adjusted or controlled according to the prediction result, an alternative power utilization method is selected, the peak-period power utilization is reduced, the power utilization cost is reduced, and the optimal allocation of power resources is realized.
Drawings
Fig. 1 is a conceptual diagram of power consumption cost saving by peak clipping and valley filling.
Fig. 2 is a flowchart of a power consumption prediction method according to embodiment 1 of the present invention.
Fig. 3 is a flow chart showing data modeling in the power consumption prediction method according to embodiment 1 of the present invention.
Fig. 4 is a flowchart of a power consumption prediction method according to embodiment 2 of the present invention.
Fig. 5 is a functional block diagram of a power consumption prediction system according to embodiment 3 of the present invention.
Detailed Description
< method for predicting Power consumption >
Example 1
Hereinafter, example 1 of the power consumption prediction method of the present invention will be described with reference to fig. 1 to 3.
Before describing the power consumption prediction method of the present invention, first, a concept of saving electricity cost by load shifting through peak clipping will be briefly described.
Fig. 1 is a graph showing a power consumption value (power consumption amount) in one day. As shown in fig. 1, when the power consumption value approaches the peak period electricity consumption interval, in order to avoid paying an expensive electricity fee, measures need to be taken to reduce the power consumption value in this interval, for example, using solar cells, photovoltaic cells, or other alternatives, or adjusting and controlling the production schedule to reduce the peak period electricity consumption.
In embodiment 1 of the present invention, a plant in which an enterprise is to execute a production plan is taken as an example of a prediction target, and power consumption to be generated when the plant executes the production plan is predicted by the power consumption prediction method of the present invention, so that power consumption in a peak period can be avoided or reduced in advance, and cost reduction can be achieved. Of course, the prediction target is not limited to a factory, and may be other power consumption units such as a mall, a hospital, and a building.
Fig. 2 is a flowchart of a power consumption prediction method according to embodiment 1 of the present invention, which includes the following steps, as shown in fig. 2.
Step S201: the factory produces according to the production plan and records the actual production data in a certain period (such as in one day), including the quantity of products produced in one day, the completion rate of the production plan in one day, the delivery rate in one day and the like;
step S202: measuring actual power consumption value data of a factory per unit time (e.g. per minute) in the day, for example, obtaining actual power consumption values per minute of each step or link through sensors installed on a production line or each production equipment, and then obtaining total actual power consumption values per minute through accumulation;
step S203: learning the recorded daily production data and the measured actual power consumption value data per unit time, and establishing a data model, which will be specifically described later;
step S204: when a factory is to perform production according to a new production plan, the new production plan is input to the data model, and the predicted power consumption value data per unit time corresponding to the new production plan is predicted.
Step S205: according to the prediction result, the power management system is controlled to actively cope with peak period power utilization or to replace it with another power utilization manner before a new production plan is formally executed.
Specifically, in step S201, when the factory performs production according to a production plan (historical production plan), data relating to production, such as the production amount (number or number of products) per product per day, the completion rate (%) of individual products per day, the progress rate (%) of the production plan per day, the shipment rate or yield per day, and the like, are recorded and stored in real time. The collected data may be stored in a storage device within the plant or may be transmitted to a dedicated server via a network. The term "one day" is merely an example, and it is needless to say that other period lengths are possible, for example, "half day" may be used as the period length for a production plan in which only half day is operated per day, or "two days" and "five days" may be used as the period length for a non-continuous operation.
On the other hand, in most cases, the production plan of a plant is largely influenced by weather or the like. Severe weather such as sunshine duration, air humidity, temperature, typhoon or heavy rain directly affects the operation of the plant parts, thereby affecting the production capacity of the day, and therefore, the actual production capacity data of each day recorded in step S201 includes other information that the weather of the day, etc. may affect the production.
In step S202, the power consumption per minute of each production device during the production process of the day is measured, the measured data is also stored in the storage device or transmitted to the server, and the total power consumption per minute during the day is obtained by summing. Here, "every minute" is merely an example, and may be every 5 minutes, every 10 minutes, or the like, and may be set as appropriate according to the accuracy requirement or the specification of the apparatus. In the present embodiment, the actual power consumption value data measured may be the actual power consumption peak value per minute of the plant, or may be an average value or the like.
In step S203, the daily production amount data and the actual power consumption value data per minute are analyzed by a control center such as a CPU or the server, and a data model is created based on the relationship between the data, as shown in fig. 3.
Fig. 3 shows a specific procedure of data model establishment based on machine learning (AI learning) in the power consumption prediction method according to embodiment 1 of the present invention.
Step S301: from the stored data of the plurality of daily production volumes and the actual power consumption values per minute, feature data is selected according to a corresponding analysis model described later, for example, features with small value change, Pearson correlation coefficients, and the like are removed. The selected daily production data includes information on the weather such as sunshine duration, air humidity, temperature, typhoon or rainstorm, and the like, as described above. For example, taking the sunshine duration as an example, the production amount data in the history period having the similar sunshine duration or sunshine intensity to the future period to be predicted is selected. The feature data may be selected by referring to a weather forecast (which may be a weather forecast broadcast by a weather station to the public or weather information predicted by a dedicated device) of a future time period to be predicted. The data of the actual power consumption values per minute includes actual power consumption value data per minute of a single product, or may include actual power consumption value data per minute of combinations (including multiple combinations) of multiple products, and the like, and is selected according to the degree of combination between the products.
Step S302: and selecting a data analysis method, and generating a corresponding analysis model aiming at the characteristic data, so that the stored daily production data and the actual power consumption value data per minute are learned to obtain the corresponding relation between the two. For example, the actual power consumption value data per minute for a single product described above may select a linear regression model, the actual power consumption value data per minute for a simple combination of a plurality of products may select a non-linear regression model, and the actual power consumption value data per minute for a complex combination of a plurality of products may select an Artificial Neural Network (ANN) model, and the like. The linear regression model, the non-linear regression model, and the artificial neural network model are merely examples of the analysis method selected in the present invention, and are not limited, and other models or custom models may be selected according to the characteristics of the stored data and the requirements of the specific analysis.
Step S303: the various analysis models generated in step S302 are verified and evaluated, for example, a plurality of sets of stored historical daily production data and actual power consumption value data per minute are substituted, and a difference between the analysis result of the analysis model and the stored data is obtained according to the corresponding relationship between the two calculated by the analysis model, so as to determine whether the current analysis model is suitable for the selected historical data. Meanwhile, the comprehensive error of the selected analysis model and the actual data is calculated by considering the influence conditions of production line idling, production product switching, production line maintenance and the like. Based on the error result, the analytical model is selected as the analytical model suitable for the selected feature data, for example, in the case where the error is less than a prescribed percentage.
Step S304: and feeding back and correcting the selected analysis model to establish a final data model. When the selected analysis model is operated, the consistency between the operation result (predicted value) of the analysis model and the corresponding actual value can be monitored in real time or at regular intervals. For example, the predicted value of the data of the power consumption value obtained by each analysis model is recorded, the predicted value is compared with the data of the actual power consumption value corresponding to the prediction, if the difference between the two values is within a predetermined range, the current analysis model is still suitable, the operation can be continued, and if the difference between the two values exceeds the predetermined range, the current analysis model deviates from an ideal state, and correction is needed, for example, various correction coefficients are calculated or interpolation is performed, so that the difference between the two values as the comparison result can return to the predetermined range, and the analysis model to be learned is always the analysis model suitable for the selected characteristic data. The analytical model thus obtained was used as the final data model. In step S304, as described above, the feedback and correction may be performed in real time, or the correction may be automatically triggered after the operation of the analysis model reaches a predetermined number of times, or the correction may be performed by an operator according to actual needs. Returning to fig. 2, after the final data model is created, in step S204, when the plant is to perform production according to a new production plan, the new production plan (daily production amount data) is input to the data model, and the predicted value of the power consumption value data per minute corresponding to the new production plan is calculated using the data model by the pairing, integration, classification of data and the AI learning and analysis described above. The predicted value can clearly indicate the actual power consumption value data per minute that will be generated when the new production plan is executed.
For example, when the production plan of the factory is mainly affected by weather conditions (other factors such as machine aging and circuit failure are not considered here), the new production plan is analyzed day by day based on the data model, and in combination with weather forecast (public information released by the weather bureau) of each day in the future or other information capable of predicting weather changes, the same or similar data is searched in the stored historical data (training data) by using the data model for pairing and machine learning, so as to obtain the power consumption peak value per minute corresponding to the new production plan.
In step S205, when the prediction result in step S204 shows a large power consumption value in the peak period power consumption interval as shown in fig. 1, the enterprise may take corresponding measures, for example, transmit the prediction value data to the power management system, and select any power utilization method such as a solar cell, a photovoltaic cell, a rechargeable battery, etc. by the power management system to replace the power utilization in the interval, thereby reducing the peak period power consumption and reducing the production cost. Moreover, the data model obtained through steps S301 to S304 is always an analysis model suitable for the selected data, thereby ensuring the accuracy and precision of the prediction.
Example 2
Although the power consumption value for executing a production plan in a large power consumption place such as a factory is predicted in embodiment 1, the power consumption prediction method of the present invention is also applicable to power consumption equipment with large power consumption, such as an air conditioner and a motor. In the present embodiment, a central air conditioner installed in a mall will be described as an example.
Fig. 4 is a flowchart of a power consumption prediction method according to embodiment 2 of the present invention, which includes the following steps, as shown in fig. 4.
Step S401: the air conditioner operates according to a preset operation plan including a daily operation period, an air volume according to a change in outside air temperature, a target temperature, an operation mode in different seasons, and the like, and records daily actual workload data of the air conditioner corresponding to these settings during actual operation, such as daily output power, a heating or cooling range, a temperature change amount due to operation of the air conditioner, and the like.
Step S402: and measuring actual power consumption value data of the air conditioner every minute in the day, namely the peak value of power consumption every minute.
Here, "every minute" is merely an example, and may be every 5 minutes, every 10 minutes, or the like, and may be set as appropriate according to the accuracy requirement or the specification of the apparatus. In the present embodiment, the measured actual power consumption value data may be the actual power consumption peak value per minute of the air conditioner, or may be an average value or the like.
Step S403: and learning the daily actual workload data and the measured actual power consumption value data per minute, and establishing a data model.
Step S404: when the air conditioner is to be operated according to a new operation plan, for example, when the dehumidification mode is operated only due to sudden temperature drop and the heating needs to be performed in advance, the corresponding operation plan is input into the data model, and the predicted power consumption value data per unit time corresponding to the new operation plan is predicted.
Step S405: and controlling the power management system according to the prediction result, and selecting other power utilization modes to replace the peak time period power utilization aiming at the high power consumption to be generated in the peak time period power utilization interval.
Other electricity utilization modes include any electricity utilization mode such as selecting a solar cell, a photovoltaic cell, a rechargeable battery and the like to replace the electricity utilization in the interval, or an enterprise appropriately adjusts the work plan so as to control the predicted value of the actual electricity consumption value corresponding to the adjusted work plan to be at the minimum in kilowatt-hour of the electricity utilization in the peak period, thereby reducing the electricity utilization in the peak period and reducing the operation cost.
The data model in step S403 adopts the same flow as the method shown in fig. 3, and therefore, the description thereof is omitted here.
By adopting the power consumption prediction method of the embodiment of the invention, for equipment with larger power consumption, such as an air conditioner, a motor and the like, the power consumption per minute can be predicted according to the working plan, so that a substitution mode or a device for peak period power utilization can be designed in advance, and the power utilization cost is reduced.
< Power consumption prediction System >
In embodiment 3, the above description has been made on the case where the power consumption prediction method of the present invention is applied to a power consumption place such as a factory and a power consumption device such as an air conditioner.
Fig. 5 is a functional block diagram of a power consumption prediction system of the present invention. The power consumption prediction system 500 includes: an acquiring module 501, configured to acquire actual production amount data or actual workload data for one day (a predetermined period) when a prediction object such as a plant or an air conditioner is produced or operated according to a production plan or an operation plan thereof, where the data may be acquired from a daily production report or a production log of the plant, an operation log of the air conditioner, or the like, or may be acquired from a control center in which various production plans or operation plans are stored; a measurement module 502 for measuring an actual power consumption peak per unit time (e.g., per minute) during a day of a factory or air conditioner; a data modeling module 503 that learns the actual production amount data or the actual workload data and the corresponding actual power consumption value data per unit time, and establishes a data model; and a prediction module 504 that inputs a new production plan to be performed by the plant or a new operation plan in which the air conditioner is to be operated to the prediction module 504, and the prediction module 504 outputs a predicted value of the power consumption peak per minute corresponding to the new production plan or the operation plan by using the final data model obtained by the data modeling module 503.
The power consumption prediction system 500 of the present invention cooperates with a power control system 510 of a plant or an air conditioner, and an actual peak value of power consumption per minute output from the prediction module 504 is input to the power control system 510, and the power control system 510 selects an alternative power consumption mode, controls various devices, or transmits an adjusted production plan scheme or the like to a production plan or a control plan maker, according to the predicted power consumption situation. Thus, when a new production plan or work plan is run, the electricity consumption cost is controlled, and green production can be realized.
The data modeling module 503 in the power consumption prediction system 500 specifically includes a feature data extracting unit 5031 for selecting feature data by screening the measured actual production data or actual workload data and the actual power consumption value data per unit time according to the process shown in fig. 3; and an AI learning unit 5032 in which the AI learning unit 5032 learns the feature data by the analysis module to generate a plurality of analysis models including a linear regression model, a nonlinear regression model, an artificial neural network model, and the like, the plurality of analysis models are verified by the verification module using the plurality of sets of measured data, the analysis model for creating the data model is selected by the selection module based on the verification result, and the analysis model is fed back and corrected by the feedback correction module. Besides the measured data of each group, the verification module can also perform verification by using the data predicted by the prediction module 504, so as to improve the accuracy of the data model and the prediction result.
The data modeling module 503 may include an input module, a storage module, a communication module, a display module, a power supply module, and the like to perform corresponding functions in addition to the above functional components. For example, the communication module communicates with an external server or a control center, and the display module displays various necessary information.
The present invention also provides a computer-readable storage medium storing a program for causing a computer (which may be a personal computer, a server, a network device, or the like) to execute the power consumption prediction method. The storage medium includes: various media capable of storing program codes, such as a U disk, a Read Only Memory (ROM), a Random Access Memory (RAM), a removable hard disk, a magnetic disk, or an optical disk.
The above-mentioned serial numbers of the embodiments of the present invention are merely for description and do not represent the merits of the embodiments.
In the above embodiments of the present invention, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments.
In the embodiments provided in the present application, it should be understood that the disclosed technology can be implemented in other ways. The above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units may be a logical division, and in actual implementation, there may be another division, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, units or modules, and may be in an electrical or other form.
The units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device to execute all or part of the steps of the method according to the embodiments of the present invention.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that, for those skilled in the art, various modifications and decorations can be made without departing from the principle of the present invention, and these modifications and decorations should also be regarded as the protection scope of the present invention.
Description of the reference symbols
500 power consumption prediction system
501 acquisition module
502 measurement module
503 data modeling module
5031 feature data extractor
5032AI learning part
504 prediction module
510 power control system.

Claims (14)

1. A power consumption prediction method, comprising the steps of:
(1) the prediction object produces or works according to the production plan or the work plan;
(2) acquiring actual production amount data or actual workload data of the prediction target in a predetermined period, and measuring actual power consumption value data per unit time, which is actual power consumption value data per unit time, of the prediction target in the predetermined period corresponding to the actual production amount data or actual workload data in the predetermined period;
(3) learning the actual production data or the actual workload data and the actual power consumption value data per unit time, and establishing a data model;
(4) when the prediction target is to perform production or work according to a new production plan or work plan, predicted power consumption value data per unit time, which is predicted power consumption value data per unit time corresponding to the new production plan or work plan, is predicted based on the data model.
2. The power consumption prediction method according to claim 1,
the predicted power consumption value data per unit time is supplied to a power management system to manage and control the actual power consumption of the prediction object.
3. The power consumption prediction method according to claim 1,
and screening the measured actual production data or actual workload data and the actual power consumption value data per unit time, and selecting characteristic data for establishing the data model.
4. The power consumption prediction method according to claim 3,
and analyzing the characteristic data to generate a plurality of analysis models including a linear regression model, a nonlinear regression model and an artificial neural network model.
5. The power consumption prediction method according to claim 4,
and substituting the measured actual production amount data or actual workload data and the actual power consumption value data per unit time into the analysis model, thereby respectively verifying and evaluating the plurality of analysis models, and selecting the analysis model for establishing the data model.
6. The power consumption prediction method according to claim 5,
feeding back and correcting the selected analysis model, establishing the data model,
and when the analysis model operates, comparing the obtained predicted power consumption value data per unit time with the actual power consumption value data per unit time corresponding to the operation, and correcting the analysis model according to a comparison result so that the comparison result is within a preset error range.
7. The power consumption prediction method according to any one of claims 1 to 6,
the actual power consumption value data per unit time and the predicted power consumption value data per unit time indicate a peak value of power consumption per unit time of the prediction target.
8. The power consumption prediction method according to any one of claims 1 to 6,
the predetermined period is one day, and the unit time is one minute.
9. The power consumption prediction method according to any one of claims 1 to 6,
the prediction object is a power consumption place including a factory, a building, a hospital, and a market.
10. The power consumption prediction method according to any one of claims 1 to 6,
the prediction target is a power consumption device including an air conditioner and a motor.
11. A power consumption prediction system, comprising:
the acquisition module is used for acquiring actual production volume data or actual workload data within a preset period when the prediction object carries out production or work according to a production plan or a work plan;
a measurement module configured to measure actual power consumption value data per unit time, which is actual power consumption value data per unit time, of the prediction target in the predetermined period, corresponding to the actual production amount data or the actual workload data in the predetermined period;
the data modeling module is used for learning the actual production data or the actual workload data and the actual power consumption value data per unit time and establishing a data model;
and a prediction module that predicts predicted power consumption value data per unit time, which is predicted power consumption value data per unit time corresponding to a new production plan or work plan, based on the data model when the prediction target is to produce or work according to the new production plan or work plan.
12. The power consumption prediction system of claim 11,
the data modeling module includes:
a feature data extraction unit that screens the measured actual production volume data or actual workload data and the actual power consumption value data per unit time, and selects feature data for creating the data model; and
and an AI learning unit which learns the feature data, generates a plurality of analysis models including a linear regression model, a nonlinear regression model, and an artificial neural network model, verifies and evaluates each of the plurality of analysis models by substituting the measured actual production amount data or actual workload data and the actual power consumption value data per unit time into the analysis model, and selects an analysis model for building the data model from the plurality of analysis models.
13. The power consumption prediction system of claim 12,
the data modeling module further comprises a feedback correction module, and the feedback correction module compares the obtained predicted power consumption value data per unit time with the actual power consumption value data per unit time corresponding to the current operation when the analysis model operates, and corrects the analysis model according to a comparison result so that the comparison result is within a preset error range.
14. A computer-readable storage medium, characterized in that,
a computer program for causing a computer to perform the method of any one of claims 1 to 10 is stored.
CN202010580966.2A 2020-06-23 2020-06-23 Power consumption prediction method, power consumption prediction system, and computer-readable storage medium Pending CN113837420A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010580966.2A CN113837420A (en) 2020-06-23 2020-06-23 Power consumption prediction method, power consumption prediction system, and computer-readable storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010580966.2A CN113837420A (en) 2020-06-23 2020-06-23 Power consumption prediction method, power consumption prediction system, and computer-readable storage medium

Publications (1)

Publication Number Publication Date
CN113837420A true CN113837420A (en) 2021-12-24

Family

ID=78963991

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010580966.2A Pending CN113837420A (en) 2020-06-23 2020-06-23 Power consumption prediction method, power consumption prediction system, and computer-readable storage medium

Country Status (1)

Country Link
CN (1) CN113837420A (en)

Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102855343A (en) * 2011-06-10 2013-01-02 阿自倍尔株式会社 Load amount prediction device, load amount prediction method, and load amount prediction program
CN103106541A (en) * 2011-11-10 2013-05-15 索尼公司 Power management device, power management method, and demand notifying device
CN104137373A (en) * 2012-02-14 2014-11-05 东芝三菱电机产业系统株式会社 Electricity demand prediction system
CN105139090A (en) * 2015-08-25 2015-12-09 国网天津市电力公司 Power industry safety stock decision analysis method based on consumption prediction
CN105787606A (en) * 2016-03-24 2016-07-20 国网辽宁省电力有限公司电力科学研究院 Power dispatching online trend early warning system based on ultra short term load prediction
CN107563539A (en) * 2017-07-24 2018-01-09 佛山市顺德区中山大学研究院 Short-term and long-medium term power load forecasting method based on machine learning model
KR101855248B1 (en) * 2017-04-20 2018-05-08 (주)에코센스 Method of energy saving performance evaluation using the truncated linear regression model
CN108053085A (en) * 2018-01-19 2018-05-18 山东大陆计量科技有限公司 Quality of production control method and device
CN109118012A (en) * 2018-08-28 2019-01-01 成都天衡智造科技有限公司 A kind of industrial dynamics various dimensions energy consumption cost prediction technique, system, storage medium and terminal
CN110414724A (en) * 2019-07-10 2019-11-05 东软集团股份有限公司 For predicting method, apparatus, readable storage medium storing program for executing and the electronic equipment of power consumption

Patent Citations (10)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102855343A (en) * 2011-06-10 2013-01-02 阿自倍尔株式会社 Load amount prediction device, load amount prediction method, and load amount prediction program
CN103106541A (en) * 2011-11-10 2013-05-15 索尼公司 Power management device, power management method, and demand notifying device
CN104137373A (en) * 2012-02-14 2014-11-05 东芝三菱电机产业系统株式会社 Electricity demand prediction system
CN105139090A (en) * 2015-08-25 2015-12-09 国网天津市电力公司 Power industry safety stock decision analysis method based on consumption prediction
CN105787606A (en) * 2016-03-24 2016-07-20 国网辽宁省电力有限公司电力科学研究院 Power dispatching online trend early warning system based on ultra short term load prediction
KR101855248B1 (en) * 2017-04-20 2018-05-08 (주)에코센스 Method of energy saving performance evaluation using the truncated linear regression model
CN107563539A (en) * 2017-07-24 2018-01-09 佛山市顺德区中山大学研究院 Short-term and long-medium term power load forecasting method based on machine learning model
CN108053085A (en) * 2018-01-19 2018-05-18 山东大陆计量科技有限公司 Quality of production control method and device
CN109118012A (en) * 2018-08-28 2019-01-01 成都天衡智造科技有限公司 A kind of industrial dynamics various dimensions energy consumption cost prediction technique, system, storage medium and terminal
CN110414724A (en) * 2019-07-10 2019-11-05 东软集团股份有限公司 For predicting method, apparatus, readable storage medium storing program for executing and the electronic equipment of power consumption

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
卢洪刚;: "高校能源消耗建模研究与应用", 环境科学与管理, no. 10 *

Similar Documents

Publication Publication Date Title
KR102187327B1 (en) Dynamic management and control system for a building electric demand based on automated machine learning scheme
KR102071377B1 (en) Fuel cell fleet optimization
EP3002848B1 (en) Demand-side grid-level load balancing aggregation system
US7951496B2 (en) Model based real-time optimization of fuel cell clusters
EP2682914A1 (en) Energy management method and system thereof, and gui method
KR20150123540A (en) A method and an apparatus operating of a smart system for optimization of power consumption
CN116683500B (en) Active power scheduling method and system for electrochemical energy storage power station
CN117217407B (en) Method, medium and device for pre-evaluating carbon emission of power system
CN117273337A (en) Intelligent electric energy meter evaluation method
CN117578534B (en) Scheduling method, device, equipment and storage medium of photovoltaic energy storage system
CN117485177A (en) Intelligent energy-saving control system and method for charging pile
CN117350507A (en) Virtual power plant scheduling system
CN117439101A (en) Intelligent network for interaction of new energy and flexible load in power grid
CN118378834A (en) Energy-saving type energy comprehensive management system and method for optimizing energy scheduling
CN117895468A (en) Cooperative control method and system for source network load storage
CN113837420A (en) Power consumption prediction method, power consumption prediction system, and computer-readable storage medium
KR20210026665A (en) Server
CN112072635A (en) Intelligent power supply and utilization system and method and intelligent power utilization system
JP2021068212A (en) Demand prediction system
CN117520797B (en) Cluster-based electric heat pump demand side response potential evaluation system
CN118336836B (en) Multi-type renewable energy peak-to-peak demand analysis method based on risk elimination
CN117249537B (en) Virtual power plant scheduling and control system and method based on central air conditioner
Frimpong et al. A model for predicting unregulated energy usage
KR102700902B1 (en) A Smart control method of dispersed photovoltaic operation system for home
KR102266098B1 (en) ESS charging/discharging shchedule producing method and system

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
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20211224

WD01 Invention patent application deemed withdrawn after publication