CN114004396A - Energy management-oriented at-home population prediction method, system, device and medium - Google Patents
Energy management-oriented at-home population prediction method, system, device and medium Download PDFInfo
- Publication number
- CN114004396A CN114004396A CN202111251603.5A CN202111251603A CN114004396A CN 114004396 A CN114004396 A CN 114004396A CN 202111251603 A CN202111251603 A CN 202111251603A CN 114004396 A CN114004396 A CN 114004396A
- Authority
- CN
- China
- Prior art keywords
- family
- data
- historical
- home
- prediction model
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000012549 training Methods 0.000 claims abstract description 13
- 230000005611 electricity Effects 0.000 claims description 33
- 230000015654 memory Effects 0.000 claims description 18
- 230000001932 seasonal effect Effects 0.000 claims description 10
- 230000010354 integration Effects 0.000 claims description 8
- 238000004590 computer program Methods 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 abstract description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 9
- 210000004027 cell Anatomy 0.000 description 7
- 238000007726 management method Methods 0.000 description 7
- 210000002569 neuron Anatomy 0.000 description 7
- 238000013135 deep learning Methods 0.000 description 6
- 238000005265 energy consumption Methods 0.000 description 5
- 230000006870 function Effects 0.000 description 5
- 230000009286 beneficial effect Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 230000006399 behavior Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000013500 data storage Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000002349 favourable effect Effects 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 229910002056 binary alloy Inorganic materials 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004134 energy conservation Methods 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR 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
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Business, Economics & Management (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Game Theory and Decision Science (AREA)
- Development Economics (AREA)
- Medical Informatics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Entrepreneurship & Innovation (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a method, a system, equipment and a medium for predicting the number of people at home facing energy management, which comprise the following steps: firstly, acquiring data and processing the data, wherein the data mainly comprises family member information, family power consumption data, member distance-from-home data, meteorological data and the like; then, constructing and training a power consumption data-based at-home people number prediction model composed of 3 layers of LSTMs and a member-to-home distance-based at-home people number prediction model composed of 4 layers of LSTMs by using the acquired data; then, a family population integrated prediction model consisting of 4 layers of DNN is constructed and trained, and finally, the hourly prediction of the family population is realized. The invention adopts an integrated learning method, synthesizes the final prediction result by utilizing the prediction results of the number of people at home of different models and different sides, and improves the precision of the prediction method compared with the prediction by using a single method.
Description
Technical Field
The invention relates to the technical field of energy management, in particular to a method, a system, equipment and a medium for predicting the number of people at home facing energy management.
Background
The household energy consumption accounts for about 30% of the total social energy consumption, the traditional household energy system adopts a extensive control mode, and no matter how many people are at home or even when no people are at home, equipment such as a water heater, a water boiler and the like operate in a mode of tracking a set value. Along with equipment such as intelligent household electrical appliances, smart jack, intelligent switch constantly get into the family, carry out energy management for the family, promote energy-conservation and safety level and provide the opportunity.
An important objective of home energy system control is user satisfaction, i.e. user comfort is met, it is obvious that whether a user is at home or how many people are at home has an important influence on the control and operation of the home energy system, taking a water heater as an example, when more people are at home, the temperature of hot water needs to be maintained at a higher temperature to meet the hot water demand of many people, and when the number of people at home is small, the temperature of hot water can be lower to avoid heat consumption, and when no people are at home, the water heater can maintain a lower water temperature and even does not burn hot water, so that the control scheme of the system is determined based on the number of people at home, and a large amount of energy can be saved. On the other hand, the considerable part of the household energy consumption is generated by the behaviors of people, the number of people at home in each period of the family is predicted, the prediction of the energy consumption of residents is facilitated, and more perfect information support is provided for intelligent power distribution and utilization at the family side.
At present, the number of people at home is not commonly researched, and the number of people at public places such as scenic spots, airports, hospitals, intersections, stations and the like is more. Generally, these methods for predicting the number of people are classified into an autoregressive model, a conventional machine learning method, and a deep learning method. These methods are currently focused mainly on the characteristics of the time series of the number of people, the characteristics and trends of social environmental factors, and further predict the number of people in the future period. The number of people in public places is the result of a large number of individual activities, the social dynamics law is hidden behind the public places, the social dynamics law has relatively stable statistical characteristics, but the number of family members is small, the randomness is higher, various influence factors related to the family members need to be considered, and the existing method for predicting the number of people in public places is difficult to follow.
Disclosure of Invention
The purpose of the invention is: the method, the system, the equipment and the medium for predicting the number of people at home facing energy management are provided, the characteristics that the deep learning can extract complex rules from a large amount of data and the method has good generalization capability can be fully utilized, the prediction information of a plurality of sides is fused to form mutual verification and supervision, and the prediction precision is favorably improved.
In order to achieve the above object, the present invention provides an energy management-oriented method for predicting the number of people at home, comprising:
acquiring historical family data and historical meteorological data of a target family, wherein the historical family data comprises: family member data, family power consumption data and family member distance data;
constructing a first at-home people number integration prediction model of the target family according to the historical family data and the historical meteorological data of the target family;
acquiring historical family number data of a target family, and inputting history family data, historical meteorological data and historical family number data of the target family into the first family number integrated prediction model for training to obtain a second family number integrated prediction model of the target family;
and acquiring real-time family data and real-time meteorological data of the target family in the time period to be predicted, and inputting the real-time family data and the real-time meteorological data into the second family population integrated prediction model to obtain the family population prediction result of the target family.
Further, the building of the first at-home people number integration prediction model of the target family according to the historical family data of the target family specifically includes:
according to the household electricity consumption data and the historical meteorological data, constructing an electricity consumption data-based family number prediction model;
according to the family member distance data and the historical meteorological data, a family number prediction model based on the member distance from the family is constructed;
and constructing a first family number integrated prediction model of the target family according to the electricity consumption data-based family number prediction model and the family number prediction model based on the distance between the members and the family.
Further, the acquisition time of the historical family data and the historical meteorological data of the target family is not less than 160 days, and the acquisition time is uniformly distributed in each season.
Further, the historical meteorological data includes: seasonal data, calendar data, and weather data, among others; the weather data includes: temperature, wind speed, humidity, and probability of rain.
The invention also provides an energy management-oriented at-home people number prediction system, which comprises: the device comprises a data acquisition module, a construction module, a training module and a prediction module, wherein the data acquisition module is used for acquiring data;
the data acquisition module is used for acquiring historical family data and historical meteorological data of a target family, wherein the historical family data comprises: family member data, family power consumption data and family member distance data;
the building module is used for building a first at-home people number integrated prediction model of the target family according to the historical family data and the historical meteorological data of the target family;
the training module is used for acquiring historical family number data of a target family, inputting the historical family number data, the historical meteorological data and the historical family number data of the target family into the first family number integrated prediction model for training, and acquiring a second family number integrated prediction model of the target family;
the prediction module is used for acquiring real-time family data and real-time meteorological data of the target family in the time period to be predicted, inputting the real-time family data and the real-time meteorological data into the second family number integrated prediction model, and obtaining the family number prediction result of the target family.
Further, the building module is specifically configured to:
according to the household electricity consumption data and the historical meteorological data, constructing an electricity consumption data-based family number prediction model;
according to the family member distance data and the historical meteorological data, a family number prediction model based on the member distance from the family is constructed;
and constructing a first family number integrated prediction model of the target family according to the electricity consumption data-based family number prediction model and the family number prediction model based on the distance between the members and the family.
Further, the acquisition time of the historical family data and the historical meteorological data of the target family is not less than 160 days, and the acquisition time is uniformly distributed in each season.
Further, the historical meteorological data includes: seasonal data, calendar data, and weather data, among others; the weather data includes: temperature, wind speed, humidity, and probability of rain.
The present invention also provides a computer terminal device, comprising: one or more processors; a memory coupled to the processor for storing one or more programs; when executed by the one or more processors, cause the one or more processors to implement the energy management oriented at-home people prediction method of any one of the above.
The invention also provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the energy management-oriented at-home people number prediction method according to any one of the above.
Compared with the prior art, the energy management-oriented method, the energy management-oriented system, the energy management-oriented device and the energy management-oriented medium for predicting the number of people at home have the beneficial effects that:
1. the invention adopts the integrated learning method, utilizes the prediction results of the number of people at home with different models and different sides to synthesize the final prediction result, and compared with the prediction by using a single method, the integrated learning method ensures the prediction method to have better precision in principle.
2. The method fully utilizes the characteristics that deep learning can extract complex rules from a large amount of data and has good generalization capability, adopts deep learning networks such as LSTM and DNN to predict the number of people at home, and is favorable for extracting the influence rule of accidental factors on the number of people at home, thereby improving the prediction precision. The LSTM can fully mine the relation between time sequences, whether family members are at home is closely related to recent behaviors, and therefore the LSTM can fully mine the time sequence characteristics and other comprehensive characteristics of the number of family members. The DNN has strong feature extraction capability, and is introduced to fuse the prediction information of a plurality of sides to form mutual check and supervision, thereby being beneficial to improving the prediction precision.
Drawings
FIG. 1 is a flow chart of a method for predicting the number of people at home facing energy management according to the present invention;
FIG. 2 is a schematic diagram of a network architecture for a model for predicting the number of people at home using electricity data provided by the present invention;
FIG. 3 is a schematic diagram of a network structure of a family population prediction model for the distance between a member and a home provided by the invention;
FIG. 4 is a schematic diagram of a network architecture of an integrated at-home people prediction model provided by the present invention;
fig. 5 is a schematic structural diagram of an energy management-oriented system for predicting the number of people at home.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be understood that the step numbers used herein are for convenience of description only and are not intended as limitations on the order in which the steps are performed.
It is to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
The terms "comprises" and "comprising" indicate the presence of the described features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The term "and/or" refers to any and all possible combinations of one or more of the associated listed items and includes such combinations.
As shown in fig. 1, the energy management-oriented method for predicting the number of people at home in the embodiment of the present invention at least includes steps S1-S4, which include the following steps:
s1, acquiring historical family data and historical meteorological data of the target family, wherein the historical family data comprises: family member data, family power consumption data and family member distance from home data.
The family member information includes: the number of family members, the number of family members going to and from work on time and the number of family students going to and from work on time, wherein the family members going to and from work on time refer to more than 80% of family members who can go to and from work on time on working days.
It should be noted that the process of acquiring the household electricity consumption data is as follows: the method comprises the steps of firstly obtaining an original data sequence of the household electric meter, then solving an average value of total household electric energy consumption in each time interval, forming a sequence according to a time sequence, and forming a household electricity consumption data sequence.
The family member off-home distance data acquisition process comprises the following steps: acquiring the linear distance between each adult member not less than 18 years old from the home of the family in each period, and arranging the data according to the following principle to form an adult member distance from home data sequence: the members are sorted according to the age from big to small, and the distance data of each member from home is arranged according to the time from first to last.
It should be noted that the historical meteorological data includes: the method comprises the following steps of obtaining seasonal data, calendar data, weather data and the like, wherein the seasonal data is which season of a target prediction time period is spring, summer, autumn and winter, the seasonal data is divided into boundary days which are spring, summer, autumn and winter in a calendar table respectively, the seasonal values are coded according to binary systems, and the spring, summer, autumn and winter are coded into 00, 01, 10 and 11 respectively.
The calendar data is information for judging whether the target prediction period is a holiday, that is: the holiday is defined as double holidays and holidays in the present application, and the target time interval is represented by the holiday 1 and is not represented by the holiday 0.
The weather data is temperature, wind speed, humidity and rainfall probability data of a target prediction time period obtained from a weather forecasting system.
S2, constructing a first at-home people number integration prediction model of the target family according to the historical family data and the historical meteorological data of the target family.
The method comprises the following specific steps:
according to the household electricity consumption data and the historical meteorological data, constructing an electricity consumption data-based family number prediction model; according to the family member distance data and the historical meteorological data, a family number prediction model based on the member distance from the family is constructed; and constructing a first family number integrated prediction model of the target family according to the electricity consumption data-based family number prediction model and the family number prediction model based on the distance between the members and the family.
In the electricity consumption data at home population prediction model, the functions of the model are as follows: for time period n, the model generates [ n +1, n +24] time period hourly number of people at home from the following input data: 1. the household electricity consumption data sequence of [ n-23, n ] time period, 2, family member information, specifically: the system comprises a family total number, a family on-time commuter number, a family commuter number, 3, season information of a time interval n, information of whether the time interval n is on a holiday or not in a time interval of 4, [ n +1, n +24], and a meteorological data sequence of the time interval n +1, n +24 in a time interval of 5, wherein the meteorological data comprise temperature, wind speed, humidity and rainfall probability data; the at-home population prediction model based on electricity consumption data is a structure consisting of 3 layers of LSTM, wherein the LSTM is an abbreviation of Long Short-term Memory and refers to a Long-Short term Memory network.
It should be noted that, as shown in fig. 2, the model is formed by connecting LSTM cells of 3 monolayers in series.
The first layer contains an LSTM cell, for time period nThe presentation period n is based on the input of the electricity consumption data at home population prediction model,representing the output of the floor i LSTM unit of the electricity usage data-based at-home population prediction model for time period n out,memory information of an i-th layer LSTM unit of the electricity consumption data-based at-home people number prediction model in a time period n is represented; then the LSTM unit of the first layer has 148 external inputs and 96 outputs going out, and for period n the LSTM unit external input isThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationPassed to time period n + 1; the above-mentioned external inputIs a one-dimensional array composed of the following data arranged in sequence: 1. [ n-23, n ]]Time interval household electricity data sequence en-23,en-22,…,enWhereinenRepresents the power consumption of the family for time period n; 2. at time interval n, the total number of family membersNumber of family members going on and off duty on timeNumber of students in family 3. Season information to which time interval n belongsSnWhereinSnRepresenting the season of the time interval n, and the value is coded according to binary, wherein the spring, summer, autumn and winter are respectively coded as 00, 01, 10 and 11; 4. [ n +1, n +24]]Information whether a time period is on a holiday hour by hourfn+1,fn+2,…,fn+24WhereinfnRepresenting whether the time interval n is on a rest day, if so, taking a value of 1 in a holiday, and otherwise, taking a value of 0; 5. [ n +1, n +24]]Temperature over a period of hoursTn+1,Tn+2,…,Tn+24Hourly wind speedwn+1,wn+2,…,wn+24Hourly humiditydn+1,dn+2,…,dn+24Hourly probability of rainfallrn+1,rn+2,…,rn+24WhereinTn、wn、dn、rnRespectively representing the temperature, the wind speed, the humidity and the rainfall probability in the time period n; namely: for period n, the external inputs to the first tier LSTM units are:
the second layer contains an LSTM unit with 96 external inputs and 48 outputs, and for time period n, the LSTM unit external input isThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationPassed to period n + 1.
The third layer contains an LSTM cellThere are 48 external inputs and 24 outputs, and for time period n, the unit external input isThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationPassed to time period n + 1; output outwardsThat is, the electricity consumption data based at-home people prediction model is output, and the at-home people prediction sequence is performed hour by hour according to the 24 hoursWhereinThe predicted value of the number of people at home in the n +1 time period given by the electricity consumption data-based number of people at home prediction model is shown.
It should be noted that, for the model for predicting the number of people at home with a distance from the family, the function of the model is that, for the period n, the model generates the number of people at home hourly in the period [ n +1, n +24] from the following input data: 1. the distance data sequence of each adult member from home in the family in the [ n-23, n ] period, 2, family member information, specifically: the system comprises a family total number, a family on-time commuter number, a family commuter number, 3, season information of a time interval n, information of whether the time interval (4) is on a holiday or not in an hour-by-hour mode in a (n +1, n + 24) time interval, and a meteorological data sequence (5) in an hour-by-hour mode in a (n +1, n + 24) time interval, wherein the meteorological data comprise temperature, wind speed, humidity and rainfall probability data; the at-home population prediction model based on the distance of a member from home is a structure consisting of 4 layers of LSTM.
It should be noted that, as shown in fig. 3, the model is formed by connecting LSTM cells of 4 monolayers in series.
The first layer contains an LSTM cell, for time period nThe presentation period n is based on the input of the at-home population prediction model of the distance of the adult from home,representing the output of the i-th layer LSTM unit of the at-home population prediction model at time period n based on the distance of the member from home,the memory information of the I-th layer LSTM unit in the time period n of the prediction model of the number of the family based on the distance between the member and the family is represented; then, the LSTM cells of the first layer have124+24NadtAn external input having 124 outputs to the outside, whereinNadtIs the number of adults in the home, and for time period n, the external input of the LSTM unit isThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationPassed to time period n + 1; [ n-23, n ]]The sequence of distances of the aged members of the time period family from the home is shown as:whereinRepresenting the distance from the family of the jth adult time period n, the members are numbered according to the natural number from big to small according to the age, and for the time period n, the external input of the first-layer LSTM unit is as follows:
the second layer contains an LSTM unit having 124 external inputs and 96 outputs, and for time period n, the LSTM unit external input isThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationTo time segment n + 1.
The third level contains an LSTM cell with 96 external inputs and 72 outputs, and for time period n, the external input of the LSTM cell isThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationPassed to period n + 1.
The fourth layer contains an LSTM unit with 72 external inputs and 24 outputs, for time period n, the external inputs beingThe outward output isAt the same time, the LSTM unit receives the memory information of the unit in the period n-1Outputting informationAnd memorize the information of the LSTM unit in the time period nOutputting informationPassed to time period n + 1; output outwardsThat is, the prediction sequence of the number of people at home 24 hours by hour in the future is based on the output of the prediction model of the number of people at home based on the distance between the members and the homeWhereinThe predicted value of the number of the people at home in the n +1 time period given by the prediction model of the number of the people at home based on the distance between the members and the home is represented.
It should be noted that, for the first at-home population integration prediction model, the function of the model is that, for the period n, the model generates [ n +1, n +24] the period of the hourly number of at-home population from the following input data: 1. the distance data sequence of each adult member from home in the family in the [ n-23, n ] period, 2, family member information, specifically: the family total number, the number of people on duty and off duty by time, the number of students in family, 3, season information of time interval n, information whether the time interval (4) is on a holiday or not in an hour-by-hour mode in [ n +1, n +24] and a meteorological data sequence (5) comprising temperature, wind speed, humidity and rainfall probability data in an hour-by-hour mode in [ n +1, n +24] time interval; the at-home population prediction model based on the distance of a member from home is a structure consisting of 4 layers of LSTM.
It should be noted that, as shown in fig. 4, the model is a feed-forward neural network with a 4-layer structure, in which 1 input layer, 2 hidden layers, and 1 output layer are arranged in the order of input layer, hidden layer 1, hidden layer 2, and output layer.
The input layer contains 169 neurons, and for the period n, the number of people is integrated for inputting the prediction modelMeaning that each neuron of the input layer is connected in sequence one-to-oneOne of the elements of (a) or (b),the composition of (A) is as follows:
the hidden layer 1 is provided with 96 neurons, the hidden layer 1 is fully connected with the input layer, the hidden layer 1 is fully connected with the hidden layer 2, and the hidden layer is not connected with other layers.
The hidden layer 2 is provided with 48 neurons, the hidden layer 2 is fully connected with the hidden layer 1, the hidden layer 2 is fully connected with the output layer, and the hidden layer is not connected with other layers.
The output layer contains 24 neurons, the output layer is fully connected with the hidden layer 2 and is not connected with other layers, each neuron of the output layer sequentially outputs one of the predicted values of the number of people at home within 24 hours in the future, and for the time period n, 24 neurons sequentially outputWhereinThe predicted value of the number of persons at home for the period n +1 generated by the integrated prediction model of the number of persons at home is represented.
S3, obtaining historical family number data of a target family, inputting the historical family number data, the historical meteorological data and the historical family number data of the target family into the first family number integrated prediction model for training, and obtaining a second family number integrated prediction model of the target family.
Specifically, the prediction model is trained by using historical data which are uniformly distributed in all seasons and are not less than 160 days, and a second at-home people number integrated prediction model is obtained.
S4, acquiring real-time family data and real-time meteorological data of the target family in the time period to be predicted, inputting the real-time family data and the real-time meteorological data into a second family population integrated prediction model, and obtaining a family population prediction result of the target family.
Specifically, a trained power consumption data-based family number prediction model, a family number prediction model based on the distance between an adult and a family, and a family number integrated prediction model are connected into a whole and put into operation, and when the family number value in the same time period is predicted for multiple times, the result predicted after use covers the result predicted first, so that the final prediction result is obtained.
In an embodiment of the present invention, the building a first at-home people number integration prediction model of the target family according to the historical family data of the target family specifically includes: according to the household electricity consumption data and the historical meteorological data, a household people number prediction model based on the electricity consumption data is built; according to the family member distance data and the historical meteorological data, constructing a family number prediction model based on the member distance from the family; and constructing a first at-home number integrated prediction model of the target family according to the at-home number prediction model based on the electricity consumption data and the at-home number prediction model based on the distance between the members and the home.
In one embodiment of the invention, the acquisition time of the historical family data and the historical meteorological data of the target family is not less than 160 days, and the acquisition time is uniformly distributed in each season.
In one embodiment of the present invention, the historical meteorological data includes: seasonal data, calendar data, and weather data, among others; the weather data includes: temperature, wind speed, humidity, and probability of rain.
Compared with the prior art, the energy management-oriented at-home people number prediction method has the beneficial effects that:
1. the invention adopts the integrated learning method, utilizes the prediction results of the number of people at home with different models and different sides to synthesize the final prediction result, and compared with the prediction by using a single method, the integrated learning method ensures the prediction method to have better precision in principle.
2. The method fully utilizes the characteristics that deep learning can extract complex rules from a large amount of data and has good generalization capability, adopts deep learning networks such as LSTM and DNN to predict the number of people at home, and is favorable for extracting the influence rule of accidental factors on the number of people at home, thereby improving the prediction precision. The LSTM can fully mine the relation between time sequences, whether family members are at home is closely related to recent behaviors, and therefore the LSTM can fully mine the time sequence characteristics and other comprehensive characteristics of the number of family members. The DNN has strong feature extraction capability, and is introduced to fuse the prediction information of a plurality of sides to form mutual check and supervision, thereby being beneficial to improving the prediction precision.
As shown in fig. 5, the present invention further provides an energy management-oriented system 200 for predicting the number of people at home, comprising: the system comprises a data acquisition module 201, a construction module 202, a training module 203 and a prediction module 204, wherein;
the data obtaining module 201 is configured to obtain historical family data and historical meteorological data of a target family, where the historical family data includes: family member data, family power consumption data and family member distance data;
the building module 202 is configured to build a first at-home population integration prediction model of the target home according to historical home data and historical meteorological data of the target home;
the training module 203 is configured to obtain historical family number data of a target family, input the historical family number data, the historical meteorological data and the historical family number data of the target family into the first family number integrated prediction model for training, and obtain a second family number integrated prediction model of the target family;
the prediction module 204 is configured to obtain real-time family data and real-time weather data of the target family within the time period to be predicted, and input the real-time family data and the real-time weather data into the second family population integration prediction model to obtain a family population prediction result of the target family.
In an embodiment of the present invention, the constructing module 202 is specifically configured to:
according to the household electricity consumption data and the historical meteorological data, constructing an electricity consumption data-based family number prediction model;
according to the family member distance data and the historical meteorological data, a family number prediction model based on the member distance from the family is constructed;
and constructing a first family number integrated prediction model of the target family according to the electricity consumption data-based family number prediction model and the family number prediction model based on the distance between the members and the family.
In one embodiment of the invention, the acquisition time of the historical family data and the historical meteorological data of the target family is not less than 160 days, and the acquisition time is uniformly distributed in each season.
In one embodiment of the present invention, the historical meteorological data includes: seasonal data, calendar data, and weather data, among others; the weather data includes: temperature, wind speed, humidity, and probability of rain.
The present invention also provides a computer terminal device, comprising: one or more processors; a memory coupled to the processor for storing one or more programs; when executed by the one or more processors, cause the one or more processors to implement the energy management oriented at-home people prediction method of any one of the above.
It should be noted that the processor may be a Central Processing Unit (CPU), other general-purpose processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), an off-the-shelf programmable gate array (FPGA) or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component, etc., the general-purpose processor may be a microprocessor, or the processor may be any conventional processor, the processor is a control center of the terminal device, and various interfaces and lines are used to connect various parts of the terminal device.
The memory mainly includes a program storage area and a data storage area, wherein the program storage area can store an operating system, an application program required by at least one function, and the like, and the data storage area can store related data and the like. In addition, the memory may be a high-speed random access memory, a non-volatile memory such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a flash memory card (FlashCard), and the like, or other volatile solid-state memory device.
It should be noted that the terminal device may include, but is not limited to, a processor and a memory, and those skilled in the art will understand that the terminal device is only an example and does not constitute a limitation of the terminal device, and may include more or less components, or combine some components, or different components.
The invention also provides a computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the energy management-oriented at-home people number prediction method according to any one of the above.
It should be noted that the computer program may be divided into one or more modules/units (e.g., computer program) which are stored in the memory and executed by the processor to implement the present invention. The one or more modules/units may be a series of computer program instruction segments capable of performing specific functions, which are used for describing the execution process of the computer program in the terminal device.
The above-mentioned embodiments are provided to further explain the objects, technical solutions and advantages of the present invention in detail, and it should be understood that the above-mentioned embodiments are only examples of the present invention and are not intended to limit the scope of the present invention. It will be understood that any modifications, equivalents, improvements and the like which come within the spirit and principle of the invention are deemed to be within the scope of the invention.
Claims (10)
1. The energy management-oriented at-home population prediction method is characterized by comprising the following steps:
acquiring historical family data and historical meteorological data of a target family, wherein the historical family data comprises: family member data, family power consumption data and family member distance data;
according to the historical family data and the historical meteorological data of the target family, a first family number integrated prediction model of the target family is constructed;
acquiring historical family number data of a target family, inputting the historical family data, the historical meteorological data and the historical family number data of the target family into the first family number integrated prediction model for training, and acquiring a second family number integrated prediction model of the target family;
and acquiring real-time family data and real-time meteorological data of the target family in the time period to be predicted, and inputting the real-time family data and the real-time meteorological data into the second family population integrated prediction model to obtain the family population prediction result of the target family.
2. The method for predicting the number of people at home according to claim 1, wherein the first integrated prediction model of the number of people at home of the target home is constructed according to historical home data of the target home, and specifically comprises:
according to the household electricity consumption data and the historical meteorological data, a household people number prediction model based on the electricity consumption data is constructed;
according to the family member distance data and the historical meteorological data, a family number prediction model based on the member distance from the family is constructed;
and constructing a first family number integrated prediction model of the target family according to the electricity consumption data-based family number prediction model and the family number prediction model based on the distance between the members and the family.
3. The method of predicting the number of people at home as claimed in claim 1, wherein the time for collecting the historical family data and the historical weather data of the target family is not less than 160 days, and the time for collecting the historical family data and the historical weather data is uniformly distributed in each season.
4. The method of predicting the number of people at home of claim 1, wherein the historical weather data comprises: seasonal data, calendar data, and weather data, among others; the weather data includes: temperature, wind speed, humidity, and probability of rain.
5. An energy management-oriented system for predicting the number of people at home, comprising: the device comprises a data acquisition module, a construction module, a training module and a prediction module, wherein the data acquisition module is used for acquiring data;
the data acquisition module is used for acquiring historical family data and historical meteorological data of a target family, wherein the historical family data comprises: family member data, family power consumption data and family member distance data;
the building module is used for building a first at-home population integration prediction model of the target family according to the historical family data and the historical meteorological data of the target family;
the training module is used for acquiring historical family number data of a target family, inputting the historical family number data, the historical meteorological data and the historical family number data of the target family into the first family number integrated prediction model for training, and acquiring a second family number integrated prediction model of the target family;
the prediction module is used for acquiring real-time family data and real-time meteorological data of the target family in the time period to be predicted, inputting the real-time family data and the real-time meteorological data into the second family number integrated prediction model, and obtaining the family number prediction result of the target family.
6. The system of claim 5, wherein the building module is configured to:
according to the household electricity consumption data and the historical meteorological data, a household people number prediction model based on the electricity consumption data is constructed;
according to the family member distance data and the historical meteorological data, a family number prediction model based on the member distance from the family is constructed;
and constructing a first family number integrated prediction model of the target family according to the electricity consumption data-based family number prediction model and the family number prediction model based on the distance between the members and the family.
7. The system of claim 5, wherein the historical family data and the historical weather data of the target family are collected for no less than 160 days, and the collection time is uniformly distributed in each season.
8. The system of claim 5, wherein the historical weather data comprises: seasonal data, calendar data, and weather data, among others; the weather data includes: temperature, wind speed, humidity, and probability of rain.
9. A computer terminal device, comprising: one or more processors; a memory coupled to the processor for storing one or more programs; when executed by the one or more processors, cause the one or more processors to implement the energy management oriented at-home population prediction method of any one of claims 1 to 4.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, implements the energy management oriented at home population prediction method according to any one of claims 1 to 4.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111251603.5A CN114004396A (en) | 2021-10-26 | 2021-10-26 | Energy management-oriented at-home population prediction method, system, device and medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111251603.5A CN114004396A (en) | 2021-10-26 | 2021-10-26 | Energy management-oriented at-home population prediction method, system, device and medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114004396A true CN114004396A (en) | 2022-02-01 |
Family
ID=79924302
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111251603.5A Pending CN114004396A (en) | 2021-10-26 | 2021-10-26 | Energy management-oriented at-home population prediction method, system, device and medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114004396A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115587904A (en) * | 2022-12-08 | 2023-01-10 | 成都秦川物联网科技股份有限公司 | Intelligent gas terminal management method, internet of things system, device and medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2017084213A (en) * | 2015-10-30 | 2017-05-18 | 三菱電機株式会社 | Information processing device and information processing method |
CN107766992A (en) * | 2017-11-09 | 2018-03-06 | 上海电力学院 | Family's daily load curve detailed predicting method based on user behavior |
JP2019036262A (en) * | 2017-08-22 | 2019-03-07 | 東京電力ホールディングス株式会社 | Family size assuming device, and device and method for assuming number of non-evacuating persons at disaster generation |
CN110807547A (en) * | 2019-10-22 | 2020-02-18 | 恒大智慧科技有限公司 | Method and system for predicting family population structure |
CN112819231A (en) * | 2021-02-05 | 2021-05-18 | 北京明略昭辉科技有限公司 | Method, device and equipment for predicting number of family users |
CN113110089A (en) * | 2021-04-29 | 2021-07-13 | 广东电网有限责任公司 | Household electrical appliance optimized use management method, device, equipment and medium |
-
2021
- 2021-10-26 CN CN202111251603.5A patent/CN114004396A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2017084213A (en) * | 2015-10-30 | 2017-05-18 | 三菱電機株式会社 | Information processing device and information processing method |
JP2019036262A (en) * | 2017-08-22 | 2019-03-07 | 東京電力ホールディングス株式会社 | Family size assuming device, and device and method for assuming number of non-evacuating persons at disaster generation |
CN107766992A (en) * | 2017-11-09 | 2018-03-06 | 上海电力学院 | Family's daily load curve detailed predicting method based on user behavior |
CN110807547A (en) * | 2019-10-22 | 2020-02-18 | 恒大智慧科技有限公司 | Method and system for predicting family population structure |
CN112819231A (en) * | 2021-02-05 | 2021-05-18 | 北京明略昭辉科技有限公司 | Method, device and equipment for predicting number of family users |
CN113110089A (en) * | 2021-04-29 | 2021-07-13 | 广东电网有限责任公司 | Household electrical appliance optimized use management method, device, equipment and medium |
Non-Patent Citations (2)
Title |
---|
包晓安;常浩浩;徐海;董亮亮;张娜;: "基于LSTM的智能家居机器学习系统预测模型研究", 浙江理工大学学报(自然科学版), no. 02, 12 December 2017 (2017-12-12), pages 224 - 231 * |
林顺富;黄娜娜;赵伦加;汤波;李东东;: "基于用户行为的家庭日负荷曲线模型", 电力建设, no. 10, 1 October 2016 (2016-10-01), pages 114 - 121 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115587904A (en) * | 2022-12-08 | 2023-01-10 | 成都秦川物联网科技股份有限公司 | Intelligent gas terminal management method, internet of things system, device and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Qing et al. | Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM | |
Afrasiabi et al. | Multi-agent microgrid energy management based on deep learning forecaster | |
Tang et al. | Short‐term power load forecasting based on multi‐layer bidirectional recurrent neural network | |
Meng et al. | Electricity price forecasting with high penetration of renewable energy using attention-based LSTM network trained by crisscross optimization | |
Wang et al. | Building load forecasting using deep neural network with efficient feature fusion | |
Hu et al. | A hybrid model based on CNN and Bi-LSTM for urban water demand prediction | |
Shukla et al. | Clustering based unit commitment with wind power uncertainty | |
CN106295899B (en) | Wind power probability density Forecasting Methodology based on genetic algorithm Yu supporting vector quantile estimate | |
Toubeau et al. | Capturing spatio-temporal dependencies in the probabilistic forecasting of distribution locational marginal prices | |
CN113141008B (en) | Data-driven power distribution network distributed new energy consumption capability assessment method | |
CN113052469B (en) | Method for calculating wind-solar-water-load complementary characteristic of small hydropower area lacking measurement runoff | |
Bagheri et al. | A novel wind power forecasting based feature selection and hybrid forecast engine bundled with honey bee mating optimization | |
Wahid et al. | Statistical features based approach (SFBA) for hourly energy consumption prediction using neural network | |
CN104463356A (en) | Photovoltaic power generation power prediction method based on multi-dimension information artificial neural network algorithm | |
Razghandi et al. | Short-term load forecasting for smart home appliances with sequence to sequence learning | |
Gong et al. | District heating systems load forecasting: a deep neural networks model based on similar day approach | |
Semero et al. | An accurate very short-term electric load forecasting model with binary genetic algorithm based feature selection for microgrid applications | |
Bai et al. | Improved Markov‐chain‐based ultra‐short‐term PV forecasting method for enhancing power system resilience | |
CN112348287A (en) | Electric power system short-term load probability density prediction method based on LSTM quantile regression | |
Zhang et al. | Interval prediction of ultra-short-term photovoltaic power based on a hybrid model | |
CN115907844A (en) | Short-term electricity price prediction method, device, equipment and medium | |
CN114004396A (en) | Energy management-oriented at-home population prediction method, system, device and medium | |
CN113344279B (en) | Resident load prediction method based on LSTM-SAM model and pooling | |
CN108694475B (en) | Short-time-scale photovoltaic cell power generation capacity prediction method based on hybrid model | |
Kim et al. | A deep-learning framework for forecasting renewable demands using variational auto-encoder and bidirectional long short-term memory |
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 |