CN117933458A - Intelligent heat supply load prediction method, intelligent heat supply load prediction system and heat supply system regulation and control method - Google Patents
Intelligent heat supply load prediction method, intelligent heat supply load prediction system and heat supply system regulation and control method Download PDFInfo
- Publication number
- CN117933458A CN117933458A CN202410054419.9A CN202410054419A CN117933458A CN 117933458 A CN117933458 A CN 117933458A CN 202410054419 A CN202410054419 A CN 202410054419A CN 117933458 A CN117933458 A CN 117933458A
- Authority
- CN
- China
- Prior art keywords
- data
- heat
- prediction
- prediction model
- load
- 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 49
- 230000005855 radiation Effects 0.000 claims abstract description 56
- 241000254158 Lampyridae Species 0.000 claims abstract description 54
- 238000010438 heat treatment Methods 0.000 claims abstract description 43
- 238000013528 artificial neural network Methods 0.000 claims abstract description 40
- 238000012544 monitoring process Methods 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 10
- 238000005457 optimization Methods 0.000 claims abstract description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 39
- 238000012549 training Methods 0.000 claims description 14
- 238000010606 normalization Methods 0.000 claims description 12
- 238000012795 verification Methods 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 8
- 238000004364 calculation method Methods 0.000 claims description 8
- 230000008569 process Effects 0.000 claims description 7
- 238000012360 testing method Methods 0.000 claims description 4
- 238000010521 absorption reaction Methods 0.000 claims description 3
- 238000010276 construction Methods 0.000 claims description 3
- 238000000556 factor analysis Methods 0.000 claims description 3
- 230000007774 longterm Effects 0.000 claims description 3
- 238000002203 pretreatment Methods 0.000 claims description 2
- 230000001105 regulatory effect Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 abstract description 3
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000010220 Pearson correlation analysis Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 238000004134 energy conservation Methods 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 230000009467 reduction Effects 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 238000012935 Averaging Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000010187 selection method Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
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
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- 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
-
- 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
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- 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
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- General Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Marketing (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Entrepreneurship & Innovation (AREA)
- General Business, Economics & Management (AREA)
- Artificial Intelligence (AREA)
- Molecular Biology (AREA)
- Tourism & Hospitality (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Game Theory and Decision Science (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Primary Health Care (AREA)
- Water Supply & Treatment (AREA)
- Educational Administration (AREA)
- Public Health (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an intelligent heat supply load prediction method, an intelligent heat supply load prediction system and a heat supply system regulation method, which comprise the following steps: collecting heat supply data and preprocessing the collected heat supply data; selecting key influencing factors of the heating load; constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm; according to the constructed multivariable gray prediction model, weather data and time data are input to obtain a solar radiation prediction result, then the solar radiation prediction result and heating power pipe network monitoring data are used as the input of a neural network prediction model optimized by a firefly algorithm, and a thermal load prediction result is output. The intelligent regulation and control system also comprises modeling of the secondary network of the heating system and intelligent regulation and control of the secondary network. According to the method, the influence of solar radiation data loss at the prediction moment on the prediction effect is considered by introducing a combined prediction method, the accuracy of heat load prediction is improved, a comprehensive and effective framework is provided for heat load prediction, and meanwhile, an intelligent regulation and control technology of a secondary network is provided, so that the method has important significance in management and optimization of a building energy system.
Description
Technical Field
The invention relates to the technical field of heat load prediction of a heat supply system, in particular to an intelligent heat load prediction method and system and a regulation and control method of the heat supply system.
Background
The statements in this section merely provide background information related to the present disclosure and may not necessarily constitute prior art.
The heat and cold supply system occupies about 5 percent of total energy consumption, and 10 to 20 percent of energy sources are wasted in the operation process. The heat supply load prediction is used as a key link for on-demand control of a heat supply system, is very important for system operation control, and has important significance for realizing building energy conservation by improving the load prediction precision.
In recent years, there are many studies on load prediction, and conventional prediction methods are based on statistics, including time series prediction, regression prediction, etc., but because there are many influencing factors of the heating load and data show highly nonlinear distribution, there is a limit in establishing an accurate mathematical model. Along with popularization of Internet of things, increasing data acquisition amount, greatly improving computing capacity and reducing informatization cost, artificial intelligence is gradually mature, machine learning is widely applied to load prediction, accuracy of a prediction model is improved due to introduction, and the problem of uncertainty of an reasoning process caused by insufficient information exists. How to be able to understand data deeply in the case of insufficient information and uncertainty is a key to achieving intelligent prediction of load.
For outdoor temperature, wind speed and relative humidity, the value of the predicted moment can be obtained according to weather forecast. Solar radiation is also an important factor affecting the heating load, but it is difficult to directly obtain a value at a predicted moment. In the prior art about heat load prediction, as in patent CN 115630561A, an automatic optimization method and apparatus for a neural network heat load prediction model are disclosed, but in the technical scheme, only temperature data and historical heat supply data for a heat supply system are obtained, but the influence of solar radiation is ignored, or in some prior art, a regression curve is adopted to obtain the value of solar radiation at the time of prediction, but the accuracy is not high, so that the prediction effect of a load model is affected by the uncertainty of an input variable. Meanwhile, based on a load prediction result, an effective heating system operation regulation technology needs to be further researched so as to realize intelligent heating in a real sense.
Disclosure of Invention
In order to solve the problems, the invention provides an intelligent heat supply load prediction method, an intelligent heat supply load prediction system and a heat supply system regulation method, considers the influence of future solar radiation on load prediction, and provides an intelligent regulation method of a secondary network based on an accurate data module.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
in a first aspect, the present invention provides a method for intelligent heating load prediction, comprising:
collecting heat supply data and preprocessing the collected heat supply data; the heat supply data comprise heat pipe network monitoring data, meteorological data and time data;
Selecting key influencing factors of solar radiation and heat supply load;
constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm;
According to the constructed multivariable gray prediction model, weather data and time data at the prediction moment are input to obtain a solar radiation prediction result, and then the solar radiation prediction result, the weather data at the prediction moment and the heat pipe network monitoring data are used as inputs of a neural network prediction model optimized by a firefly algorithm to output a heat load prediction result.
Further technical solutions, the time data includes date, day and time; the meteorological data comprises outdoor temperature, wind speed, relative humidity, solar radiation, total cloud cover and weather conditions; the heating power pipe network monitoring data comprise a water supply and return temperature, accumulated heat and accumulated flow; the accumulated flow difference between the front moment and the rear moment is used for obtaining the circulation flow, and the accumulated heat difference between the front moment and the rear moment is used for obtaining the time-by-time heat, namely the heat load;
The pretreatment method comprises the following steps: for missing data, using interpolation, namely filling the average value of the front and rear moments, and if the front and rear moment data are missing, adopting the average value of the corresponding moments of two adjacent days;
For abnormal data, using interpolation method to replace the data; and (3) deleting and repeating long-term data caused by abnormal acquisition process, and directly eliminating the data.
According to a further technical scheme, the selection of the key influence factors of solar radiation comprises the following steps: days, time, total cloud cover, and weather conditions; the key influence factors of the selected heating load comprise: the temperature of the supplied water, the heat load, the circulation flow at the previous moment, the outdoor temperature, the wind speed, the relative humidity and the solar radiation at the next moment.
According to a further technical scheme, the steps of constructing the multivariable gray prediction model are as follows:
And calculating the association degree by using an association coefficient definition formula, determining the input and output variables of the multi-variable gray prediction model, and forming a data set, wherein the association coefficient definition formula is as follows:
the association degree calculation formula is as follows:
and constructing a multivariable gray prediction model by utilizing the data set, inputting N-1 meteorological and time data as related factor sequences, and inputting solar radiation as an original sequence to obtain the multivariable gray prediction model.
According to a further technical scheme, the step of the firefly algorithm optimized neural network prediction model is as follows:
dividing training data and verification data according to monitoring data and meteorological data of a heating power pipe network, wherein the training data are used for network training, and the verification data are used for testing fitting performance of a network;
Carrying out data normalization; the input and output data is normalized by using a minimum and maximum normalization method, and the calculation formula is as follows:
Wherein x is an initial value before data normalization, y is a calculated value after data normalization, and x max and x min are maximum and minimum values obtained by original data respectively;
setting parameters and determining the number of nodes; the number of hidden layer nodes of the neural network prediction model is n, an optimal neural network structure is obtained, and a calculation formula of the hidden layer nodes is as follows: wherein a is an input layer variable, b is an output layer variable, m is an integer and 0< m <10;
Initializing firefly parameters, including the number of fireflies, step factors, light intensity absorption coefficients, maximum attraction coefficients and maximum iteration times;
starting firefly optimization to obtain an optimal weight threshold corresponding to an optimal firefly individual, and further obtaining a neural network prediction model optimized by an initial firefly algorithm;
Inputting training data into a neural network prediction model optimized by an initial firefly algorithm to obtain a trained prediction model; and verifying the trained prediction model through verification data, so as to obtain a final neural network prediction model optimized by a firefly algorithm.
According to a further technical scheme, the heat supply load is predicted, and the specific method comprises the following steps of:
inputting the time, the number of days, the total cloud quantity and the weather condition corresponding to the predicted solar radiation data into a multivariable gray prediction model to obtain a solar radiation prediction result at the next moment;
And (3) sending the water supply and return temperature, the heat load and the circulating flow at the previous moment, and the outdoor temperature, the wind speed, the relative humidity and the solar radiation at the next moment into a neural network prediction model optimized by a firefly algorithm to obtain the heat load at the next moment.
In a second aspect, the present invention provides an intelligent load prediction system comprising:
The data acquisition and preprocessing module is configured to: collecting heat supply data and preprocessing the collected heat supply data; the heat supply data comprise heat pipe network monitoring data, meteorological data and time data;
an influence factor analysis module configured to: selecting key influencing factors of solar radiation and heat load;
A predictive model construction module configured to: constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm;
A thermal load prediction module. Is configured to: according to the constructed multivariable gray prediction model, weather data and time data at the prediction moment are input to obtain a solar radiation prediction result, and then the solar radiation prediction result, the weather data at the prediction moment and the heat pipe network monitoring data are used as inputs of a neural network prediction model optimized by a firefly algorithm to output a heat load prediction result.
In a third aspect, the present invention provides a smart load control method, based on the smart load prediction method as described in the first aspect, including: the modeling of the secondary network of the heating system and intelligent regulation and control of the secondary network are carried out, and the modeling of the secondary network of the heating system comprises a heat user module, a secondary pipe network module and a heat exchanger module.
According to a further technical scheme, the hot user module is as follows: Wherein Q t,i is the predicted value of heat load of the heat input level, GJ/h; q t-1,i -the heat load of the first 1h, GJ/h; t 2g(t-1),i -water supply temperature 1h before the ith heating power inlet of the secondary network, and the temperature is lower than the temperature; t 2h(t-1),i -backwater temperature of 1h before the ith heating power inlet of the secondary network, and the temperature is DEG C; q 2g(t-1),i -circulation flow of 1h before the ith heating power inlet of the secondary network, m 3/h;Tt -outdoor temperature forecast value, DEG C; w t -outdoor wind speed forecast value, m/s; h t -outdoor relative humidity forecast,%; s t, an outdoor solar radiation predicted value, W/m 2.h;
The secondary pipe network module is as follows: Wherein, t i -ith heating power user regulates and controls the target room temperature and the temperature; setting the temperature to be 22 ℃ generally, and setting the water supply flow of q 2g -secondary network to be m 3/h;
the heat exchanger module is: Wherein q 1,q2g is the flow rate of the primary and secondary networks, m 3/h;t1g(t-1),t1h(t-1) is the temperature of the primary network water supply return in the past 1 hour; t 2g -water supply temperature of the secondary network, and the temperature is lower than the temperature.
Further technical scheme, the wisdom regulation and control of secondary net specifically includes:
The heat load of each heating power inlet at the next moment is obtained by utilizing a heat user module;
Determining the water supply flow and the water supply temperature of each heat inlet at the next moment by utilizing a secondary pipe network module;
determining the circulating water flow rate of the primary network at the next moment by using a heat exchanger module;
And executing the regulation signals, and respectively transmitting the regulation signals of the water supply flow of each heat inlet and the return water flow of the primary network at the next moment to different regulating valves, so as to realize intelligent regulation of the secondary network.
Compared with the prior art, the invention has the beneficial effects that:
1. The influence of future solar radiation on the heat load is considered, and the accuracy of the prediction model is improved;
2. The prediction method of ELMAN neural network optimized by combining the multivariable gray prediction model with the firefly algorithm is adopted, wherein the multivariable gray prediction model is adopted to mine the intrinsic law of data by utilizing the differential equation, a large number of data samples are not needed, the short-term prediction effect is good, and meanwhile, the weighting threshold of the ELMAN neural network structure is optimized by utilizing the firefly algorithm, so that the load prediction precision is improved to a certain extent;
3. Through the deep integration of the heating system and the intelligent regulation technology, the on-demand heating of the secondary network is truly realized, and the aims of energy conservation and emission reduction are achieved.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention.
FIG. 1 is a flow chart of a smart heating load prediction method and a control method of a heating system according to the present invention;
FIG. 2 is a flowchart of a ELMAN neural network algorithm optimized by the firefly algorithm of the present invention;
FIG. 3 is a flow chart of the intelligent regulation technique of the secondary network of the present invention.
The specific embodiment is as follows:
the invention is further described below with reference to the drawings and examples.
It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the invention. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
Embodiments of the invention and features of the embodiments may be combined with each other without conflict.
Example 1
As shown in fig. 1, the embodiment provides an intelligent heat supply load prediction method, which includes the steps of:
S1: collecting heat supply data and preprocessing the collected heat supply data; the heat supply data comprise heat pipe network monitoring data, meteorological data and time data;
S2: selecting key influencing factors of solar radiation and heat load;
S3: constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm;
S4: according to the constructed multivariable gray prediction model, weather data and time data at the prediction moment are input to obtain a solar radiation prediction result, and then the solar radiation prediction result, the weather data at the prediction moment and the heat pipe network monitoring data are used as inputs of a neural network prediction model optimized by a firefly algorithm to output a heat load prediction result.
In step S1, heat supply data is collected and the collected heat supply data is preprocessed, which specifically includes: the whole weather integral point data of a whole heating season in a certain place is derived from a weather platform, the derived heat supply data comprise weather data and time data, and the specific data types are as follows: reference place name, day, date, time, outdoor temperature, wind speed, relative humidity, solar radiation, total cloud cover, weather conditions, etc.;
Besides, the required heat supply data also comprises heat supply pipe network monitoring data, the data are usually from measuring signals of sensors and executing signals of actuators, heat supply operation data of each heat supply inlet of a certain district 2021 in the whole heating season are collected through an intelligent heat supply visual integrated management platform, the collecting time interval is 3 minutes for enhancing the accuracy of the data, and the data types comprise water supply and return temperature, accumulated heat quantity, accumulated flow quantity and the like; then, the derived data file is stored as xlsx files, a program is written by using Python, time-by-time screening is carried out on the data of the xlsx files with the aim of deriving time data closest to each whole point moment, heat supply system operation data at approximately the whole point moment is obtained, and time-by-time heat and circulation flow are obtained by carrying out difference calculation on accumulated heat and flow at the front moment and the rear moment; wherein the time-by-time heat is the heat load.
Then, because the operation mechanism of the energy system is complex, various problems easily occur in the process of data acquisition and storage, the whole quality of the data is low, and the missing value and the abnormal value exist, so that the data needs to be preprocessed; the pretreatment specifically comprises the following steps: for missing data, interpolation is used for filling. The filling method comprises the following steps: filling the average value of the front and rear moments, and adopting the average value of the corresponding moments of two adjacent days if the data of the front and rear moments are missing; for abnormal data, using a 3Sigma criterion to identify abnormal values and then using interpolation to replace the data; for long-term data missing and repeated data caused by abnormal acquisition, such data are directly rejected.
In step S2, a heat load key influencing factor is selected through time-by-time meteorological data and heat pipe network monitoring data, and the selection method is Pearson correlation analysis, specifically: the thermal load key influencing factors selected by using the Pearson correlation analysis comprise: the water supply temperature, the backwater temperature, the heat load, the circulation flow at the previous moment, the outdoor temperature, the wind speed, the relative humidity and the solar radiation at the next moment.
In step S3, a multivariate gray prediction model OGM (1, n) is constructed, specifically:
and collecting solar radiation data, actual measurement data of meteorological parameters and time data at m moments. The measured data of the meteorological parameters include: total cloud cover, weather conditions, etc., the time data includes: days, dates and times. The test period is selected to be 48h or more, the time interval is 1h, and each item of data corresponding to the same time is taken as an influence component. Firstly, marking a solar radiation sequence as X 0, collecting a plurality of influencing factors as X 1,X2 and …, carrying out some pretreatment on data because the units of the factors are not uniform, carrying out dimensionless treatment by adopting averaging treatment, marking the pretreated solar radiation sequence as Y 0, marking the related factor sequence as Y 1,Y2 and …, defining the correlation coefficient as formula (1), taking rho as a resolution coefficient and taking 0.5,
Wherein ρ is a resolution coefficient, 0.5, i is a variable, k is the kth element of the variable sequence,Generating a sequence for accumulation of solar radiation,/>Generating a kth element corresponding to the sequence for accumulation formed by the ith variable;
the formula of the association degree is defined as formula (2):
where n is the total number of raw data contained by each variable, k is the kth element of the variable sequence,
And (3) according to the relevancy ranking of the factors, determining the number of days, time, total cloud quantity, weather conditions and solar radiation intensity as input and output variables of a multivariable gray prediction model, and forming an m multiplied by5 data set a.
Constructing a multivariable gray prediction model OGM (1, N) by utilizing the data set a, inputting N-1 meteorological data and time data as related factor sequences, and inputting solar radiation as an original sequence to obtain a gray prediction model corresponding formula OGM X1 (1) (k). The modeling concrete process is as follows:
The system is provided with a characteristic data sequence:
X1 (0)＝[x1 (0)(1),x1 (0)(2),…,x1 (0)(n)] (3)
Related factor sequence:
X2 (0)＝[x2 (0)(1),x2 (0)(2),…,x2 (0)(n)] (4-1)
X3 (0)＝[x3 (0)(1),x3 (0)(2),…,x3 (0)(n)] (4-2)
…
XN (0)＝[xN (0)(1),xN (0)(2),…,xN (0)(n)] (4-3)
Let X i (0) (i=1, 2, … N) have the 1-AGO sequence X i (1), wherein,
X i (1) is generated next to the mean sequence Z 1 (1), wherein,
Therefore, the formula (5) is OGM (1, N):
wherein, in the OGM (1, n) model, a is called a development coefficient, b i is called a driving coefficient, and b ixi (1) (k) is called a driving term.
Β= (a, b 1,b2,…bN)T), β= (β Tβ)-1βT Y, when the variation amplitude of X i (1) (i=1, 2, … N) is small, the OGM (1, N) approximation time corresponds to the formula:
The predictive result obtained by the subtraction reduction formula is:
OGM (1, N) differential analog is:
Performing residual error check to verify model reliability, and obtaining predicted value according to the above The residual sequence is shown as formula (10), the relative error is shown as formula (11), and the average relative error is generally required to be controlled within 0.5%.
The method comprises the steps of constructing a ELMAN neural network prediction model optimized by a firefly algorithm, as shown in fig. 2, specifically:
Firstly, according to monitoring data and meteorological data of a heating power pipe network, dividing training data and verification data, wherein 90% of data sets are used as training data for network training, and 10% of data sets are used as verification data for testing fitting performance of a network; then, data normalization processing: the input and output data are normalized by using a minimum and maximum normalization method, a calculation formula is shown as formula (12), the data are normalized by using formula (12), and the data are calculated to be values of the [ -1,1] interval.
Wherein x is an initial value before data normalization, y is a calculated value after data normalization, and x max and x min are maximum and minimum values obtained from the original data respectively.
Then, setting parameters, determining the number of nodes, determining the number n of hidden layer nodes of the ELMAN neural network prediction model according to an empirical formula and a large amount of experimental experience, and obtaining the optimal ELMAN neural network structure, wherein the number of hidden layer nodes is calculated as shown in a formula (13).
Wherein a is an input layer variable, b is an output layer variable, m is an integer and 0< m <10.
Then, initializing firefly parameters including firefly number, step factor, light intensity absorption coefficient, maximum attraction coefficient and maximum iteration number, and in this embodiment, determining ELMAN a neural network structure as follows: 8-10-1, the maximum iteration number is set to be 1000, and the learning rate is set to be 0.01.
Then, starting firefly optimization to obtain an optimal weight threshold corresponding to the optimal firefly individual: a. the initial population is randomly generated. Randomly generating initial positions of fireflies in a solution space according to the variable range; b. calculating the brightness value of each firefly, and determining the moving direction of the fireflies according to the brightness value, wherein the brighter fireflies attract the surrounding darker fireflies to move to the fireflies; c. updating the firefly position; d. calculating the brightness of fireflies after updating the positions; repeating the steps b, c and d until a preset stopping condition is reached, and outputting an optimal solution. In this embodiment, the determined optimal weight threshold includes: the weight theta before the hidden layer is input, the weight gamma from the hidden layer to the receiving layer, the weight omega from the receiving layer to the output layer, the neuron threshold u of the hidden layer and the neuron threshold v of the output layer.
Finally, inputting training data into a neural network prediction model optimized by an initial firefly algorithm to obtain a trained prediction model; and verifying the trained prediction model through verification data, so as to obtain a final neural network prediction model optimized by a firefly algorithm, and unfolding the prediction.
In step S4, the heat load is predicted based on the two prediction models, which specifically includes:
Firstly, sending the time, the number of days, the total cloud quantity and the weather condition corresponding to the predicted solar radiation data to an OGM X1 (1) (k) corresponding to a gray prediction model to obtain a solar radiation prediction result at the next moment;
Then, load data, water supply temperature, backwater temperature and circulating flow of each heat inlet are extracted from a heat pipe network monitoring data set, corresponding outdoor temperature, wind speed, relative humidity and solar radiation are extracted from meteorological data in a concentrated mode, a training set and a verification set are constructed, wherein the water supply temperature, backwater temperature, heat load and circulating flow at the moment t-1, the outdoor temperature, wind speed, relative humidity and solar radiation at the moment t are input, and the heat load at the moment t is output;
Then, the training data are sent into a predicted model of ELMAN algorithm optimized based on firefly algorithm to obtain a trained predicted model, and the trained predicted model is verified through verification data to obtain a predicted model;
Then, the water supply temperature, the backwater temperature, the heat load, the circulating flow, the outdoor temperature, the wind speed, the relative humidity and the solar radiation at the next moment in the past 1h are sent into a prediction model of ELMAN algorithm optimized based on firefly algorithm, and the heat load prediction result at the next moment is obtained.
Example two
The embodiment provides an intelligent load prediction system, which specifically comprises the following modules:
The data acquisition and preprocessing module is configured to: collecting heat supply data and preprocessing the collected heat supply data; the heat supply data comprise heat pipe network monitoring data, meteorological data and time data;
an influence factor analysis module configured to: selecting key influencing factors of solar radiation and heat load;
A predictive model construction module configured to: constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm;
A thermal load prediction module. Is configured to: according to the constructed multivariable gray prediction model, weather data and time data at the prediction moment are input to obtain a solar radiation prediction result, and then the solar radiation prediction result, the weather data at the prediction moment and the heat pipe network monitoring data are used as inputs of a neural network prediction model optimized by a firefly algorithm to output a heat load prediction result.
It should be noted that, each module in the embodiment corresponds to the method in the first embodiment one by one, and the implementation process is the same, which is not described here again.
Example III
The present embodiment provides an intelligent load control method, which is based on the intelligent load prediction method mentioned in the first embodiment, and includes: modeling and intelligent regulation of a secondary network of the heating system are carried out, and the modeling of the secondary network of the heating system comprises a heat user module, a secondary pipe network module and a heat exchanger module.
First, the hot user module:
(16)
Wherein Q t,i is the predicted value of heat load of the heat input level, GJ/h;
Q t-1,i -the heat load of the first 1h, GJ/h;
t 2g(t-1),i -water supply temperature 1h before the ith heating power inlet of the secondary network, and the temperature is lower than the temperature;
T 2h(t-1),i -backwater temperature of 1h before the ith heating power inlet of the secondary network, and the temperature is DEG C;
q 2g(t-1),i -circulation flow 1h before the ith heat input port of the secondary network, m 3/h;
t t -an outdoor temperature forecast value, DEG C;
W t -outdoor wind speed forecast value, m/s;
h t -outdoor relative humidity forecast,%;
s t -predicted value of outdoor solar radiation, W/m 2.h.
Then, the secondary pipe network module:
Wherein, t i -ith heating power user regulates and controls the target room temperature and the temperature; typically set at 22 ℃.
Q 2g -secondary network water supply flow, m 3/h, and solving the total secondary network flow q 2g by establishing a fitting curve of a secondary network circulating water pump with respect to frequency-flow.
Finally, the heat exchanger module:
(18)
Wherein q 1,q2g is the flow of the primary and secondary networks, m 3/h;
t 1g(t-1),t1h(t-1) -the temperature of the primary net water supply and return in the past 1 hour, and the temperature is lower than the temperature;
t 2g -water supply temperature of the secondary network, and the temperature is lower than the temperature.
The intelligent regulation of the secondary network, as shown in fig. 3, specifically includes:
firstly, predicting the load of each heat inlet at the next moment by utilizing a ELMAN neural network heat user model optimized by a firefly algorithm;
Then, determining the water supply flow and the water supply temperature of each heating power inlet at the next moment through a secondary network model;
then, determining the primary network circulating water flow at the next moment by utilizing a ELMAN neural network heat exchanger model optimized by a firefly algorithm;
and finally, executing regulation signals, and respectively transmitting the regulation signals of the water supply flow of each heating power inlet and the return water flow of the primary network at the next moment to different regulation valves.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
While the foregoing description of the embodiments of the present invention has been presented in conjunction with the drawings, it should be understood that it is not intended to limit the scope of the invention, but rather, it is intended to cover all modifications or variations within the scope of the invention as defined by the claims of the present invention.
Claims (10)
1. An intelligent heating load prediction method, comprising:
collecting heat supply data and preprocessing the collected heat supply data; the heat supply data comprise heat pipe network monitoring data, meteorological data and time data;
Selecting key influencing factors of solar radiation and heat supply load;
constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm;
According to the constructed multivariable gray prediction model, weather data and time data at the prediction moment are input to obtain a solar radiation prediction result, and then the solar radiation prediction result, the weather data at the prediction moment and the heat pipe network monitoring data are used as inputs of a neural network prediction model optimized by a firefly algorithm to output a heat load prediction result.
2. The intelligent load prediction method according to claim 1, wherein the time data includes date and time; the meteorological data comprises outdoor temperature, wind speed, relative humidity, solar radiation, cloud cover and weather conditions; the heating power pipe network monitoring data comprise a water supply and return temperature, accumulated heat and accumulated flow; the circulation flow is obtained by making difference between the accumulated flow at the front and rear moments, and the time-by-time heat, namely the heat load, is obtained by making difference between the accumulated heat at the front and rear moments.
The pretreatment method comprises the following steps: for missing data, using interpolation, namely filling the average value of the front and rear moments, and if the front and rear moment data are missing, adopting the average value of the corresponding moments of two adjacent days;
For abnormal data, using interpolation method to replace the data; and (3) deleting and repeating long-term data caused by abnormal acquisition process, and directly eliminating the data.
3. The intelligent load prediction method according to claim 2, wherein the selecting solar radiation key influencing factors comprises: days, time, total cloud cover, and weather conditions; the key influence factors of the selected heating load comprise: the temperature of the supplied water, the heat load, the circulation flow at the previous moment, the outdoor temperature, the wind speed, the relative humidity and the solar radiation at the next moment.
4. The intelligent load prediction method according to claim 2, wherein the step of constructing the multivariate gray prediction model comprises:
And calculating the association degree by using an association coefficient definition formula, determining the input and output variables of the multi-variable gray prediction model, and forming a data set, wherein the association coefficient definition formula is as follows:
the association degree calculation formula is as follows:
and constructing a multivariable gray prediction model by utilizing the data set, inputting N-1 meteorological and time data as related factor sequences, and inputting solar radiation as an original sequence to obtain the multivariable gray prediction model.
5. The intelligent load prediction method as claimed in claim 1, wherein the step of the firefly algorithm-optimized neural network prediction model is as follows:
dividing training data and verification data according to monitoring data and meteorological data of a heating power pipe network, wherein the training data are used for network training, and the verification data are used for testing fitting performance of a network;
Carrying out data normalization; the input and output data is normalized by using a minimum and maximum normalization method, and the calculation formula is as follows:
Wherein x is an initial value before data normalization, y is a calculated value after data normalization, and x max and x min are maximum and minimum values obtained by original data respectively;
setting parameters and determining the number of nodes; the number of hidden layer nodes of the neural network prediction model is n, an optimal neural network structure is obtained, and a calculation formula of the hidden layer nodes is as follows: wherein a is an input layer variable, b is an output layer variable, m is an integer and 0< m <10;
Initializing firefly parameters, including the number of fireflies, step factors, light intensity absorption coefficients, maximum attraction coefficients and maximum iteration times;
starting firefly optimization to obtain an optimal weight threshold corresponding to an optimal firefly individual, and further obtaining a neural network prediction model optimized by an initial firefly algorithm;
Inputting training data into a neural network prediction model optimized by an initial firefly algorithm to obtain a trained prediction model; and verifying the trained prediction model through verification data, so as to obtain a final neural network prediction model optimized by a firefly algorithm.
6. The intelligent load prediction method according to claim 1, wherein the heat supply load is predicted by:
inputting the time, the number of days, the total cloud quantity and the weather condition corresponding to the predicted solar radiation data into a multivariable gray prediction model to obtain a solar radiation prediction result at the next moment;
And (3) sending the water supply and return temperature, the heat load and the circulating flow at the previous moment, the outdoor temperature, the wind speed, the relative humidity and the solar radiation at the next moment into a neural network prediction model optimized by a firefly algorithm to obtain a heat load prediction result at the next moment.
7. An intelligent load prediction system, comprising:
The data acquisition and preprocessing module is configured to: collecting heat supply data and preprocessing the collected heat supply data; the heat supply data comprise heat pipe network monitoring data, meteorological data and time data;
an influence factor analysis module configured to: selecting key influencing factors of solar radiation and heat load;
A predictive model construction module configured to: constructing a multivariable gray prediction model and a neural network prediction model optimized by a firefly algorithm;
A thermal load prediction module. Is configured to: according to the constructed multivariable gray prediction model, weather data and time data at the prediction moment are input to obtain a solar radiation prediction result, and then the solar radiation prediction result, the weather data at the prediction moment and the heat pipe network monitoring data are used as inputs of a neural network prediction model optimized by a firefly algorithm to output a heat load prediction result.
8. A method for intelligent load regulation, based on the intelligent load prediction method according to any one of claims 1-6, characterized in that the modeling of the secondary network of the heating system and the intelligent regulation of the secondary network, said modeling of the secondary network of the heating system comprises a heat user module, a secondary pipe network module and a heat exchanger module.
9. The intelligent load regulation method of claim 8 wherein the thermal user module is: Wherein Q t,i is the predicted value of heat load of the heat input level, GJ/h; q t-1,i -the heat load of the first 1h, GJ/h; t 2g(t-1),i -water supply temperature 1h before the ith heating power inlet of the secondary network, and the temperature is lower than the temperature; t 2h(t-1),i -backwater temperature of 1h before the ith heating power inlet of the secondary network, and the temperature is DEG C; q 2g(t-1),i -circulation flow of 1h before the ith heating power inlet of the secondary network, m 3/h;Tt -outdoor temperature forecast value, DEG C; w t -outdoor wind speed forecast value, m/s; h t -outdoor relative humidity forecast,%; s t, an outdoor solar radiation predicted value, W/m 2.h;
The secondary pipe network module is as follows: Wherein, t i -ith heating power user regulates and controls the target room temperature and the temperature; setting the temperature to be 22 ℃ generally, and setting the water supply flow of q 2g -secondary network to be m 3/h;
the heat exchanger module is: Wherein q 1,q2g is the flow rate of the primary and secondary networks, m 3/h;t1g(t-1),t1h(t-1) is the temperature of the primary network water supply return in the past 1 hour; t 2g -water supply temperature of the secondary network, and the temperature is lower than the temperature.
10. The intelligent load control method according to claim 8, wherein the intelligent control of the secondary network specifically comprises:
The heat load of each heating power inlet at the next moment is obtained by utilizing a heat user module;
Determining the water supply flow and the water supply temperature of each heat inlet at the next moment by utilizing a secondary pipe network module;
determining the circulating water flow rate of the primary network at the next moment by using a heat exchanger module;
And executing the regulation signals, and respectively transmitting the regulation signals of the water supply flow of each heat inlet and the return water flow of the primary network at the next moment to different regulating valves, so as to realize intelligent regulation of the secondary network.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410054419.9A CN117933458A (en) | 2024-01-12 | 2024-01-12 | Intelligent heat supply load prediction method, intelligent heat supply load prediction system and heat supply system regulation and control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410054419.9A CN117933458A (en) | 2024-01-12 | 2024-01-12 | Intelligent heat supply load prediction method, intelligent heat supply load prediction system and heat supply system regulation and control method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117933458A true CN117933458A (en) | 2024-04-26 |
Family
ID=90753079
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410054419.9A Pending CN117933458A (en) | 2024-01-12 | 2024-01-12 | Intelligent heat supply load prediction method, intelligent heat supply load prediction system and heat supply system regulation and control method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117933458A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118469349A (en) * | 2024-07-11 | 2024-08-09 | 北京中能北方科技股份有限公司 | Intelligent heat supply monitoring data analysis method and data processing equipment |
-
2024
- 2024-01-12 CN CN202410054419.9A patent/CN117933458A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118469349A (en) * | 2024-07-11 | 2024-08-09 | 北京中能北方科技股份有限公司 | Intelligent heat supply monitoring data analysis method and data processing equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gao et al. | Day-ahead power forecasting in a large-scale photovoltaic plant based on weather classification using LSTM | |
| CN109284870A (en) | Short-term method for forecasting photovoltaic power generation quantity based on shot and long term Memory Neural Networks | |
| CN108921339B (en) | Quantile regression-based photovoltaic power interval prediction method for genetic support vector machine | |
| CN116187601B (en) | Comprehensive energy system operation optimization method based on load prediction | |
| CN110046743A (en) | Energy Consumption of Public Buildings prediction technique and system based on GA-ANN | |
| CN109948845A (en) | A kind of distribution network load shot and long term Memory Neural Networks prediction technique | |
| CN117933458A (en) | Intelligent heat supply load prediction method, intelligent heat supply load prediction system and heat supply system regulation and control method | |
| CN113554466A (en) | Short-term power consumption prediction model construction method, prediction method and device | |
| CN113537582B (en) | Photovoltaic power ultra-short-term prediction method based on short-wave radiation correction | |
| CN116526473A (en) | Particle swarm optimization LSTM-based electrothermal load prediction method | |
| CN115965132A (en) | Distributed photovoltaic digital twin system power prediction method based on GA-BP neural network | |
| CN110598929A (en) | Wind power nonparametric probability interval ultrashort term prediction method | |
| Dong et al. | Short-term building cooling load prediction model based on DwdAdam-ILSTM algorithm: A case study of a commercial building | |
| CN104820877A (en) | Photovoltaic system generation power prediction method based on cloud adaptive PSO-SNN | |
| CN106446440A (en) | Short-term photovoltaic generation power prediction method based on online sequential extreme learning machine | |
| CN108376294A (en) | A kind of heat load prediction method of energy supply feedback and meteorologic factor | |
| CN114757104A (en) | Construction method of series gate group water transfer engineering hydraulic real-time regulation model based on data driving | |
| CN113762387A (en) | Data center station multi-load prediction method based on hybrid model prediction | |
| CN115115125A (en) | Photovoltaic power interval probability prediction method based on deep learning fusion model | |
| Boersma et al. | Robust sample-based model predictive control of a greenhouse system with parametric uncertainty | |
| Shi et al. | Building heating load forecasting based on the theory of transient heat transfer and deep learning | |
| CN113222231A (en) | Wisdom heating system based on internet of things | |
| CN113112085A (en) | New energy station power generation load prediction method based on BP neural network | |
| CN117454168A (en) | Building energy consumption prediction method and system | |
| Matushkin et al. | Analysis of factors for forecasting electric power generation by solar power plants |
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 |