CN116955963A - Heating ventilation energy-saving ladder optimizing control method based on historical data analysis - Google Patents
Heating ventilation energy-saving ladder optimizing control method based on historical data analysis Download PDFInfo
- Publication number
- CN116955963A CN116955963A CN202311204306.4A CN202311204306A CN116955963A CN 116955963 A CN116955963 A CN 116955963A CN 202311204306 A CN202311204306 A CN 202311204306A CN 116955963 A CN116955963 A CN 116955963A
- Authority
- CN
- China
- Prior art keywords
- data
- historical data
- distance
- historical
- sets
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 72
- 238000010438 heat treatment Methods 0.000 title claims abstract description 65
- 238000009423 ventilation Methods 0.000 title claims abstract description 65
- 238000007405 data analysis Methods 0.000 title claims abstract description 19
- 238000004378 air conditioning Methods 0.000 claims abstract description 34
- 238000011156 evaluation Methods 0.000 claims abstract description 24
- 238000005457 optimization Methods 0.000 claims description 30
- 238000005057 refrigeration Methods 0.000 claims description 24
- 238000001816 cooling Methods 0.000 claims description 23
- 230000008569 process Effects 0.000 claims description 20
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 17
- 230000002787 reinforcement Effects 0.000 claims description 10
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000004146 energy storage Methods 0.000 claims description 6
- 239000000178 monomer Substances 0.000 claims description 6
- 238000007710 freezing Methods 0.000 claims description 4
- 230000008014 freezing Effects 0.000 claims description 4
- 239000008400 supply water Substances 0.000 claims description 3
- 238000004134 energy conservation Methods 0.000 abstract description 4
- 238000005265 energy consumption Methods 0.000 description 8
- 239000004973 liquid crystal related substance Substances 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000010219 correlation analysis Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013433 optimization analysis Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention relates to the technical field of heating ventilation energy conservation, and provides a heating ventilation energy conservation ladder optimizing control method based on historical data analysis, which comprises the following steps: acquiring a plurality of historical data sets and the historical data sets at the same time according to the acquired data; acquiring a data category association degree adjustment coefficient; acquiring data distances among different historical data, further acquiring a b adjacent data distance and a b+1 adjacent data distance corresponding to the historical data, further acquiring a b adjacent difference, and further determining a first threshold; acquiring a density data distance and determining the density distance; acquiring the evaluation distance between every two simultaneous historical data sets; the method comprises the steps of constructing a heating ventilation air conditioning system model and a model of each equipment of the heating ventilation air conditioning system, obtaining optimal control parameters of the heating ventilation air conditioning system according to collected data, the established model and an evaluation distance, realizing ladder optimizing control, and solving the problem that the existing heating ventilation energy-saving control method can not enable different equipment of the heating ventilation air conditioning to reach an optimal state when running.
Description
Technical Field
The invention relates to the technical field of heating ventilation energy conservation, in particular to a heating ventilation energy conservation ladder optimizing control method based on historical data analysis.
Background
The heating and ventilation system is a system with the largest energy consumption ratio in a building, and provides comfortable and healthy indoor environment for building users by consuming energy. In order to meet the increasing demands of users on the comfort of the internal environment of a building, the energy consumption of a heating ventilation air conditioning system is gradually increased, the input cost of the heating ventilation air conditioning in the use process and the negative influence on the environment are increased, and therefore the energy consumption of the heating ventilation air conditioning needs to be controlled on the premise of ensuring the comfort of the indoor environment. The aim is achieved by improving the running state of the refrigeration system of the data center, eliminating the problem of uncooled cooperation among devices of the traditional refrigeration system, reducing the labor input cost and reducing the power consumption of the refrigeration system.
At present, the control method of the running state of the central refrigerating system of the heating ventilation air conditioner is mainly divided into a control method based on a model and a control method without a model. The model-based control method is widely researched and verified by students, but the existing heating ventilation air conditioning model can only control the whole air conditioning parameters according to heating ventilation air conditioning data, when the energy consumption of certain equipment is too high, the parameters of other equipment are often adjusted to parameters corresponding to the high energy consumption of the equipment in order to realize the whole balance of the heating ventilation air conditioning due to the limitation of the model, and the optimal operation state of a heating ventilation air conditioning central refrigerating system cannot be achieved. Therefore, a heating ventilation energy-saving ladder optimizing control method capable of realizing accurate control of each device in heating ventilation air conditioner is needed.
Disclosure of Invention
The invention provides a heating ventilation energy-saving ladder optimizing control method based on historical data analysis, which aims to solve the problem that the existing heating ventilation energy-saving control method can not enable different equipment of heating ventilation air conditioner to reach an optimal state in operation, and the adopted technical scheme is as follows:
the embodiment of the invention provides a heating ventilation energy-saving ladder optimizing control method based on historical data analysis, which comprises the following steps of:
setting a site server to collect data and transmitting the data to an AI algorithm system, so as to achieve the purpose of acquiring a plurality of historical data sets and the historical data sets at the same time according to the collected data;
acquiring the relative relevancy of data types among the historical data sets, and acquiring a data type relevancy adjustment coefficient according to the relative relevancy of the data types among the historical data sets;
acquiring data distances among different historical data according to all the historical data contained in the historical data set, and further acquiring data corresponding to the historical dataProximity data distanceThe distance between adjacent data is further obtained to obtain the corresponding first data of the historical dataAdjacent difference, for the firstClustering the adjacent differences to obtain a clustering result, and obtaining the first clustering resultA threshold value;
acquiring a density data distance corresponding to the historical data according to the data distance of the historical data and a first threshold value, and determining the density distance between two historical data corresponding to the same data type in the historical data groups at two same moments according to the density data distance corresponding to the historical data contained in the historical data groups;
acquiring the evaluation distance between every two simultaneous historical data sets;
and obtaining optimal control parameters of each device of the heating ventilation air conditioner by using an AI algorithm system according to the evaluation distance between the historical data sets at the same time, so as to realize the step optimizing control.
Further, the setting field server collects data and transmits the data to the AI algorithm system, so as to achieve the purpose of obtaining a plurality of historical data sets and the historical data sets at the same time according to the collected data, and the specific method comprises the following steps:
the site server is used for collecting data and transmitting the collected data to the AI algorithm system;
recording the collected data corresponding to the previous moment as historical data;
recording the collected data at the current moment as first-level data, and analyzing the first-level data and the historical data together as the historical data;
recording a data group consisting of historical data of the same data type as a historical data group;
and recording a data set consisting of all data needing to be acquired in the same data acquisition time as a historical data set at the same time.
Further, the method for obtaining the data category relative association degree between the historical data sets and obtaining the data category association degree adjustment coefficient according to the data category relative association degree between the historical data sets comprises the following specific steps:
using association degree analysis to all the history data sets at the same time in each history data set to obtain association coefficients between every two different types of history data;
taking each historical data set as a master data set, and recording the historical data set which is not the master data set as a slave data set;
the ratio of the gray correlation coefficient between the slave data set and the corresponding master data set to the sum of the gray correlation coefficients between the master data and all the corresponding slave data sets is recorded as a first ratio;
the ratio of the gray correlation coefficient between the slave data set and the corresponding master data set and the sum of the gray correlation coefficients between all the master data sets and the corresponding slave data sets is recorded as a second ratio;
the product of the first ratio and the second ratio is recorded as the relative relativity of the data types between the slave data set and the corresponding master data set;
and (3) recording the ratio of the data type relative association degree to the sum of the data type relative association degrees between all the slave data sets and the corresponding master data sets as a data type association degree adjustment coefficient between the slave data sets and the corresponding master data sets.
Further, the data distance between different historical data is obtained according to all the historical data contained in the historical data group, so as to obtain the corresponding historical dataProximity data distanceThe distance between adjacent data is further obtained to obtain the corresponding first data of the historical dataThe approach difference comprises the following specific methods:
recording absolute values of differences between every two historical data in the same historical data set as data distances corresponding to the two historical data;
ordering all data distances corresponding to the historical data from small to large, obtaining an ordering sequence, and taking the first data distance in the ordering sequenceDistance of data, will beNumber of piecesCorresponding to the distance recorded as history dataA proximity data distance;
obtaining historical data corresponding toA proximity data distance;
acquiring data corresponding to history dataA first preset threshold data distance closest to the data distance is recorded as a standard deviation corresponding to the historical dataA first standard deviation of the proximity data distance;
corresponding historical dataProximity data distanceThe difference value of the adjacent data distance is marked as a first difference value;
marking the sum of the first difference value and the second preset threshold value as a first sum value;
corresponding to historical dataThe product of the first standard deviation of the adjacent data distance and the first sum is recorded as the first corresponding to the historical dataThe proximity difference.
Further, the pair ofClustering is carried out on the adjacent differences to obtain a clustering result, and a first threshold value is obtained according to the clustering result, wherein the specific method comprises the following steps:
correspond the historical data toThe proximity difference is marked as the proximity difference corresponding to the historical data;
density clustering is used for adjacent differences corresponding to each historical data, and a plurality of clusters are obtained;
the average value of the adjacent differences contained in each cluster is recorded as a cluster average value, the adjacent difference closest to the cluster average value in the adjacent differences contained in the cluster with the smallest cluster average value is selected, and the selected adjacent difference closest to the cluster average value corresponds toThe value of (2) is recorded as a first threshold.
Further, the specific method for acquiring the density data distance corresponding to the historical data according to the data distance of the historical data and the first threshold value includes:
and respectively taking each historical data in the historical data set as the historical data to be analyzed, selecting a first threshold historical data with the smallest data distance with the historical data to be analyzed in the historical data set, and recording the sum of the data distances of the selected first threshold historical data value and the historical data to be analyzed as the density data distance corresponding to the historical data to be analyzed.
Further, the determining the density distance between two pieces of history data corresponding to the same data category in two simultaneous history data sets according to the density data distance corresponding to the history data included in the history data sets includes the following specific methods:
recording absolute values of differences of density data distances corresponding to two historical data corresponding to the same data type in the two simultaneous historical data sets as first absolute values;
the first absolute value is noted as the density distance.
Further, the specific method for acquiring the evaluation distance comprises the following steps:
in the method, in the process of the invention,for simultaneous historical data setsAnd time history data setThe evaluation distance between the two;for simultaneous historical data setsAnd time history data setCorresponding to the same data categoryDensity distance between two historic data of (c), wherein,；the number of different data types contained in the historical data set at the same moment;data typesAnd data categoryAnd (5) adjusting the coefficient of the data category association degree.
Further, each equipment of the heating ventilation air conditioner comprises a field control cabinet, a cooling tower, a cooling pump, a water chilling unit, a freezing pump and a plate heat exchanger.
Further, the method for obtaining the optimal control parameters of each device of the heating ventilation air conditioner by using the AI algorithm system according to the evaluation distance between the time history data sets to realize the step optimizing control comprises the following specific steps:
the step optimizing control process of the heating ventilation air conditioner comprises 5 stages, namely L1-L5, the L1-L5 process is gradually optimized and analyzed through reinforcement learning by using an AI algorithm system, and after analysis results are implemented, feedback results are fed back to the AI algorithm system, so that continuous updating iteration and control optimization are realized;
the L1 stage is used for carrying out monomer optimization on all equipment of the heating ventilation air conditioner, including but not limited to cooling tower power optimization of a cold source system, water pump frequency optimization, chiller power optimization, control logic of a precise air conditioner and optimization of fan rotating speed;
the L2 stage is used for carrying out combined optimization on all the monomer optimizing devices of the L1, and specifically comprises group control optimizing under a single refrigeration unit and under a combined and coordinated working condition of all the devices and group control optimizing of a precise air conditioning unit;
the L3 stage is used for performing joint debugging on the single refrigeration unit in the L2 stage, and optimizing among the multiple refrigeration units, including but not limited to optimizing the starting quantity of the units of the cold source control system, optimizing the unit load distribution and controlling and optimizing the round-robin work among the multiple refrigeration units;
the L4 stage is used for optimizing a plurality of sets of refrigeration units in the L3 and optimizing the tail end precise air conditioner in a linkage manner, and determining optimal chilled water supply water temperature and tail end air quantity parameters;
and the L5 stage is used for carrying out full-system optimization on the IT equipment air-conditioning system, the movable ring system, the refrigeration system and the energy storage system by carrying out full-equipment linkage on the IT equipment air-conditioning system, the movable ring system, the refrigeration system and the energy storage system.
The beneficial effects of the invention are as follows: according to the invention, through analyzing the historical data and the current time data of the operation of the heating ventilation air conditioner, the similarity and the relevance of the attributes among different types of data are considered, the data type relevance adjustment coefficient between the slave data set and the corresponding master data set is obtained, and the data type relevance adjustment coefficient between the slave data set and the corresponding master data set is obtained; then, measuring the distance between the history data sets at the same time, taking the numerical value difference between the data values as a basis, and simultaneously taking the distribution density between the history data into consideration, so as to measure the distance between the history data sets at the same time more accurately, specifically, acquiring the b-th adjacent difference corresponding to each history data according to the data density distribution condition of each history data in the corresponding data type, clustering the b-th adjacent difference corresponding to each history data to determine a first threshold value, acquiring the density data distance corresponding to the history data to be analyzed, and determining the density distance between two history data corresponding to the same data type in the two history data sets at the same time according to the first threshold value corresponding to each history data set and the density data distance corresponding to the history data to be analyzed; then, according to the data type association degree adjustment coefficient and the density distance, determining an evaluation distance between two simultaneous historical data sets, wherein the evaluation distance acquiring process comprehensively considers the similarity and the association of attributes among different types of data, and simultaneously takes the numerical difference among data values as the basis, and considers the distribution density among the historical data, so that the distance among the simultaneous historical data sets is measured more accurately, and a more accurate basis is provided for improving the control accuracy of the running state of the air conditioner; and finally, establishing a model of the heating ventilation air conditioning system and a model of each device of the heating ventilation air conditioning system, obtaining the optimal control parameters corresponding to each device in the heating ventilation air conditioning system according to the established model and the evaluation distance between the historical data sets of the same time, and realizing the operation of the whole system in an optimal energy-saving mode by a step-type optimizing control method, thereby solving the problem that the existing heating ventilation air conditioning model can only control the whole heating ventilation air conditioning system and is easy to cause high energy consumption of air conditioning operation.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
Fig. 1 is a schematic flow chart of a heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to an embodiment of the invention.
Fig. 2 is a schematic diagram of a ladder optimization system.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, a flowchart of a heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to an embodiment of the invention is shown, and the method includes the following steps.
And S001, setting a site server to collect data and transmitting the data to an AI algorithm system, so as to achieve the purpose of acquiring a plurality of historical data sets and the historical data sets at the same time according to the collected data.
In order to realize the step optimizing control of the heating ventilation air conditioner, a heating ventilation air conditioner step optimizing control system is established, and the system comprises an AI algorithm system, a field server, a field control cabinet, a cooling tower, a cooling pump, a water chilling unit, a freezing pump and a plate heat exchanger, wherein the numbers of the equipment corresponding to FIG. 2 are respectively 1-8, and the interrelated modes are shown in FIG. 2.
The site server is used for collecting data and transmitting the collected data to the AI algorithm system; the data to be collected by the field server comprises water inlet and outlet temperature of a cooling tower, water flow of the cooling tower, fan frequency of the cooling tower, outdoor temperature and humidity, frequency and flow of a cooling pump, refrigerating capacity of a water chilling unit, refrigerating side inlet and outlet temperature of the water chilling unit, electric load of the water chilling unit, and frequency and flow of the refrigerating pump.
The processing method of the acquired data comprises the following steps:
in order to facilitate subsequent analysis, the data corresponding to the acquired previous moment is recorded as historical data, the acquired data at the current moment is marked as first-level data, and the first-level data and the historical data are used together as the historical data for analysis. Meanwhile, the effect of marking the collected data at the current moment as first-level data is as follows: and distinguishing the data corresponding to the acquired previous time is facilitated.
And acquiring a plurality of historical data sets and a simultaneous historical data set according to the historical data of the heating ventilation air conditioner. The historical data set is a data set formed by the obtained historical data with the same data type, for example, all the water inlet and outlet temperatures of the cooling towers in the obtained historical data form a set of historical data set, and the data type corresponding to the historical data set is the water inlet and outlet temperatures of the cooling towers. The historical data set at the same time is a data set formed by all the historical data which are acquired at the same data acquisition time and need to acquire data types. The time interval for the field server to collect the historical data can be set according to the needs of the implementer.
To this end, a plurality of history data sets and a simultaneous history data set are acquired.
Step S002, obtaining the relative relevance of the data types among the historical data sets, and obtaining the data type relevance adjustment coefficient according to the relative relevance of the data types among the historical data sets.
An AI algorithm is used to determine an evaluation distance between each two simultaneous histories from the histories and the simultaneous histories.
In the existing method for controlling the heating, ventilation and air conditioning according to the historical data, the difference between the data acquired at different moments is only measured by the numerical difference, and the similarity and the relevance of the attributes among the data are not considered, so that when the energy consumption of some equipment is too high, the parameters of other equipment can be adjusted to the parameters corresponding to the higher energy consumption of the equipment in order to realize the overall balance of the heating, ventilation and air conditioning, and the problem that the operation of a central refrigerating system of the heating, ventilation and air conditioning tends to be in an optimal state cannot be solved.
And analyzing all the historical data sets at the same time in each historical data set by using a gray correlation analysis GRA algorithm to acquire gray correlation coefficients between every two different types of historical data. When the correlation between two different kinds of history data is larger, the gray correlation coefficient between the two different kinds of history data is larger.
And respectively taking each historical data set as a historical data set to be analyzed, recording the historical data set to be analyzed as a master data set, recording the historical data set which is not the master data set as a slave data set, and acquiring a data type association degree adjustment coefficient between the slave data set and the corresponding master data set when each historical data set is taken as the master data set.
In the method, in the process of the invention,for slave data setsCorresponding main data groupThe relative relativity of the data types between the two;for slave data setsCorresponding main data groupGray correlation coefficients between;is the main data groupAll slave data sets corresponding theretoThe sum of gray correlation coefficients between the two;the sum of gray correlation coefficients between all master data sets and all slave data sets corresponding to the master data sets is used.
When the correlation between the slave data set and the corresponding master data set is larger, the gray correlation coefficient between the two data sets is larger, the relative correlation degree of the data types between the slave data set and the corresponding master data set is larger, namely the correlation between the slave data set and the data types acquired by the corresponding master data set according to the historical data is larger.
And acquiring a data category correlation degree adjustment coefficient between the slave data set and the corresponding master data set according to the data category relative correlation degree between the slave data set and the corresponding master data set.
In the method, in the process of the invention,for slave data setsCorresponding main data groupThe data category association degree adjustment coefficient between the two;for slave data setsCorresponding main data groupThe relative relativity of the data types between the two;for all slave data sets and their corresponding master data setsThe sum of the relative relatedness of the data types between them.
And acquiring the data type relevance adjustment coefficient between the two data sets according to the data type relative relevance between the data sets and the corresponding main data sets, wherein the data type relevance adjustment coefficient is used for adjusting the value range of the relevance evaluation between the two data sets, and when the data type relative relevance between the data sets and the corresponding main data sets is larger, the data type relevance adjustment coefficient between the two data sets is larger, namely the relevance between the data sets acquired from the data sets and the corresponding main data sets according to the historical data is larger.
So far, the data category association degree adjustment coefficient between the historical data sets is obtained.
Step S003, obtaining the data distance between different historical data according to all the historical data contained in the historical data group, and further obtaining the corresponding historical dataProximity data distanceThe distance between adjacent data is further obtained to obtain the corresponding first data of the historical dataAdjacent difference, for the firstClustering is carried out on the adjacent differences to obtain a clustering result, and a first threshold is obtained according to the clustering result.
After considering the similarity and the relevance of the attributes between the data, the distance between the historical data sets at the same time is measured, and the conventional distance measuring method directly measures the distance according to the numerical difference between the data values, so that the distance measurement is more accurate, the distance between the historical data sets at the same time is measured more accurately by taking the numerical difference between the data values as a basis and considering the distribution density between the historical data.
All the historical data contained in the same historical data set are analyzed, and the distribution density among the historical data is evaluated. Respectively taking the same history dataAnd recording the absolute value of the difference value between every two historical data in the group as the data distance corresponding to the two historical data. Each historical data in the historical data group corresponds to a plurality of data distances, all the data distances corresponding to the historical data are ordered from small to large, and the first data distance is selectedDistance of data, will beThe data distance is recorded as the corresponding historical dataAdjacent data distance.
Similarly, obtain the corresponding historical dataAdjacent data distance.
Acquisition and acquisition ofThe proximity data is closest in distanceData distance, this will beThe standard deviation of the data distance is recorded asA first standard deviation of the proximity data distance. Wherein, the liquid crystal display device comprises a liquid crystal display device,，the number of data distances corresponding to the historical data;for the first constant, the empirical value is 8, and the practitioner can take the value of the first constant according to the requirementSetting.
When the numerical value of the history data is distributed at a position with larger distribution density in the history data group corresponding to the history data, the numerical value of the history data is distributed at a position with larger distribution densityThe smaller the first standard deviation of the adjacent data distance,Proximity data distanceThe smaller the proximity data distance difference.
Acquiring the first corresponding to each historical data according to the analysisThe proximity difference.
In the method, in the process of the invention,for historical data setsInternal history dataCorresponding firstA proximity difference;is historical dataCorresponding toA first standard deviation of the proximity data distance;is historical dataCorresponding toA proximity data distance;is historical dataCorresponding toA proximity data distance;is a second constant.
The experience value of the second constant is 1, and an operator can set the value of the second constant according to the requirement, so that the second constant has the function of preventing the situation that the adjacent difference cannot accurately evaluate the data distribution density corresponding to the historical data due to the fact that the value of the second constant is 0 in an absolute value symbol.
Correspond the historical data toThe adjacent difference is marked as the adjacent difference corresponding to the historical data, so that each historical data can obtain the corresponding historical dataThe proximity differences.
Corresponding to each history dataClustering the adjacent differences by using a DBSCAN clustering algorithm with Epsilon=5 and minPts=4 to obtain a plurality of clusters, marking the average value of the adjacent differences contained in each cluster as a cluster average value, taking the adjacent difference closest to the cluster average value from the adjacent differences contained in the cluster with the smallest cluster average value, and corresponding the adjacent differenceThe value of (2) is recorded as a first threshold value. The first threshold is a threshold which is determined according to the density condition of the data distribution in the historical data set and is used for counting the range of the data distribution density around each data.
Thus, a first threshold is obtained.
Step S004, acquiring a density data distance corresponding to the historical data according to the data distance of the historical data and the first threshold value, and determining the density distance between two historical data corresponding to the same data category in the historical data sets at two same moments according to the density data distance corresponding to the historical data contained in the historical data sets.
And respectively taking each historical data in the historical data set as the historical data to be analyzed, selecting a first threshold historical data with the smallest data distance with the historical data to be analyzed in the historical data set, calculating the sum of the data distances between the selected first threshold historical data value and the historical data to be analyzed, and recording the sum as the density data distance corresponding to the historical data to be analyzed.
And determining the density distance between two historical data corresponding to the same data category in the two simultaneous historical data sets according to the first threshold value corresponding to each historical data set.
In the method, in the process of the invention,for simultaneous historical data setsInternal history dataAnd time history data setInternal history dataDensity distance between, wherein, historical dataWith historical dataCorresponds to the same data type;for history data set at same timeInternal history dataA corresponding density data distance;for history data set at same timeInternal history dataCorresponding density data distance.
When the density data distance difference between the two historical data is smaller, the density distance between the two historical data is smaller, and the two simultaneous historical data sets corresponding to the two historical data are closer and more likely to correspond to the historical data corresponding to the heating ventilation air conditioner in the same working state.
Thus, the density distance between two historical data corresponding to the same data type in the two simultaneous historical data sets is obtained.
Step S005, obtaining the evaluation distance between every two simultaneous historic data sets.
And determining the evaluation distance between the two simultaneous historical data sets according to the data type association degree adjustment coefficient between the slave data set and the corresponding master data set and the density distance between the two historical data corresponding to the same data type in the two simultaneous historical data sets.
In the method, in the process of the invention,for simultaneous historical data setsAnd time history data setThe evaluation distance between the two;for simultaneous historical data setsAnd time history data setCorresponding to the same data categoryDensity distance between two historic data of (c), wherein,；the number of different data types contained in the historical data set at the same moment;data typesAnd data categoryAnd (5) adjusting the coefficient of the data category association degree.
When the density distance between two historical data corresponding to the same data type in the two simultaneous historical data sets is larger, the evaluation distance between the two simultaneous historical data sets is larger, and the two simultaneous historical data sets are more likely to correspond to data corresponding to heating ventilation and air conditioning under different working states.
So far, the evaluation distances among different time historic data sets are obtained.
And S006, obtaining optimal control parameters of each device of the heating ventilation air conditioner by using an AI algorithm system according to the evaluation distance between the historical data sets at the same time, and realizing step optimizing control.
The AI algorithm system adopts a deep learning algorithm to acquire optimized control parameters of each device and the whole system, the optimized control parameters are transmitted to a field control cabinet through a field server, and the field control cabinet transmits the optimized control parameters to a cooling tower, a cooling pump, a water chilling unit, a freezing pump and a plate heat exchanger, so that the operation of the whole system in an optimal energy-saving mode is realized.
In particular, the AI algorithm system realizes the digital calculation and control of the physical model through reinforcement learning, adopts big data analysis and judgment to obtain the optimal solution of each device and the whole system, and optimally controls the field device in a shortcut mode.
The reinforcement learning is a known technique, and will not be described in detail.
Specifically, the step optimizing control process of the heating ventilation air conditioner comprises 5 stages, namely L1-L5, the AI algorithm system gradually performs optimizing analysis through the L1-L5 process, and the analysis result is fed back to the AI algorithm system after implementation, so that continuous updating iteration and control optimization are realized. Wherein, the optimization analysis process of L1-L5 is realized by using reinforcement learning.
In particular, the L1 stage is used for performing monomer optimization on all equipment of the heating ventilation air conditioner, and comprises cooling tower power optimization, water pump frequency optimization, chiller power optimization, control logic of the precise air conditioner and optimization of fan rotating speed.
Taking optimizing the power of the cooling tower of the cold source system as an example, the optimizing of the power of the cooling tower of the cold source system is realized by reinforcement learning, and the deep Q reinforcement learning model is taken as an example, and the model is input as the evaluation distance between all time history data sets corresponding to the cooling tower and output as the power of the cooling tower. The deep Q reinforcement learning model is known in the prior art, and training is required to be performed on the model in the process of using the deep Q reinforcement learning model, and the training process is known to those skilled in the art and will not be repeated.
In particular, the L2 stage is used for carrying out combined optimization on all the monomer optimizing devices of the L1, and specifically comprises group control optimizing under a single refrigeration unit and under a combined and coordinated working condition of all the devices and group control optimizing of a precise air conditioning unit.
Further, the method specifically comprises the combined optimization of a single refrigeration unit lower chiller, a cooling tower, a refrigeration pump, a cooling pump and a plate heat exchanger, and the combined optimization of a tail end precise air conditioning unit.
In particular, the L3 stage is used for performing joint debugging on a single set of refrigeration units in the L2 stage, and optimizing among a plurality of sets of refrigeration units, including optimizing the starting quantity of units of a cold source control system, optimizing unit load distribution and optimizing round-robin work control among the plurality of sets of refrigeration units.
In particular, the L4 stage is used for optimizing a plurality of sets of refrigeration units in the L3 and the terminal precise air conditioner in a linked manner, and determining the optimal chilled water supply water temperature and terminal air quantity parameters.
In particular, the L5 stage is used for performing full-system optimization on the IT equipment air conditioning system, the moving ring system, the refrigerating system and the energy storage system by performing full-equipment linkage on the IT equipment air conditioning system, the moving ring system, the refrigerating system and the energy storage system.
The optimization in the L2-L5 stages adopts deep reinforcement learning to realize the optimization control of the air conditioning system, and the process of realizing the optimization control is the prior known technology and is not repeated.
It should be noted that the sequence of the embodiments of the present invention is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims can be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (10)
1. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis is characterized by comprising the following steps of:
setting a site server to collect data and transmitting the data to an AI algorithm system, so as to achieve the purpose of acquiring a plurality of historical data sets and the historical data sets at the same time according to the collected data;
acquiring the relative relevancy of data types among the historical data sets, and acquiring a data type relevancy adjustment coefficient according to the relative relevancy of the data types among the historical data sets;
acquiring data distances among different historical data according to all the historical data contained in the historical data set, and further acquiring data corresponding to the historical dataProximity data distance and->The distance between adjacent data is further obtained, and the corresponding +.>Adjacent difference, for the->Proximity differenceClustering is carried out to obtain a clustering result, and a first threshold value is obtained according to the clustering result;
acquiring a density data distance corresponding to the historical data according to the data distance of the historical data and a first threshold value, and determining the density distance between two historical data corresponding to the same data type in the historical data groups at two same moments according to the density data distance corresponding to the historical data contained in the historical data groups;
acquiring the evaluation distance between every two simultaneous historical data sets;
and obtaining optimal control parameters of each device of the heating ventilation air conditioner by using an AI algorithm system according to the evaluation distance between the historical data sets at the same time, so as to realize the step optimizing control.
2. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the setting field server collects data and transmits the data to an AI algorithm system, the purpose of obtaining a plurality of historical data sets and the historical data sets at the same time according to the collected data is achieved, the specific method comprises the following steps:
the site server is used for collecting data and transmitting the collected data to the AI algorithm system;
recording the collected data corresponding to the previous moment as historical data;
recording the collected data at the current moment as first-level data, and analyzing the first-level data and the historical data together as the historical data;
recording a data group consisting of historical data of the same data type as a historical data group;
and recording a data set consisting of all data needing to be acquired in the same data acquisition time as a historical data set at the same time.
3. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the obtaining the data type relative association between the historical data sets and obtaining the data type association adjustment coefficient according to the data type relative association between the historical data sets comprises the following specific steps:
using association degree analysis to all the history data sets at the same time in each history data set to obtain association coefficients between every two different types of history data;
taking each historical data set as a master data set, and recording the historical data set which is not the master data set as a slave data set;
the ratio of the gray correlation coefficient between the slave data set and the corresponding master data set to the sum of the gray correlation coefficients between the master data and all the corresponding slave data sets is recorded as a first ratio;
the ratio of the gray correlation coefficient between the slave data set and the corresponding master data set and the sum of the gray correlation coefficients between all the master data sets and the corresponding slave data sets is recorded as a second ratio;
the product of the first ratio and the second ratio is recorded as the relative relativity of the data types between the slave data set and the corresponding master data set;
and (3) recording the ratio of the data type relative association degree to the sum of the data type relative association degrees between all the slave data sets and the corresponding master data sets as a data type association degree adjustment coefficient between the slave data sets and the corresponding master data sets.
4. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the data distance between different historical data is obtained according to all the historical data contained in the historical data group, and then the corresponding historical data is obtainedProximity data distance and->The distance between adjacent data is further obtained, and the corresponding +.>Proximity toThe difference comprises the following specific methods:
recording absolute values of differences between every two historical data in the same historical data set as data distances corresponding to the two historical data;
ordering all data distances corresponding to the historical data from small to large, obtaining an ordering sequence, and taking the first data distance in the ordering sequenceDistance of data, will be->The data distance is recorded as +.>A proximity data distance;
obtaining historical data corresponding toA proximity data distance;
acquiring data corresponding to history dataA first preset threshold data distance closest to the data distance is recorded as the standard deviation of the first preset threshold data distance corresponding to the historical data>A first standard deviation of the proximity data distance;
corresponding historical dataProximity data distance and->The difference value of the adjacent data distance is marked as a first difference value;
marking the sum of the first difference value and the second preset threshold value as a first sum value;
corresponding to historical dataThe product of the first standard deviation of the adjacent data distance and the first sum is recorded as the +.>The proximity difference.
5. The history data analysis-based heating, ventilation and energy saving ladder optimizing control method according to claim 1, wherein the pair of first stepsClustering is carried out on the adjacent differences to obtain a clustering result, and a first threshold value is obtained according to the clustering result, wherein the specific method comprises the following steps:
correspond the historical data toThe proximity difference is marked as the proximity difference corresponding to the historical data;
density clustering is used for adjacent differences corresponding to each historical data, and a plurality of clusters are obtained;
the average value of the adjacent differences contained in each cluster is recorded as a cluster average value, the adjacent difference closest to the cluster average value in the adjacent differences contained in the cluster with the smallest cluster average value is selected, and the selected adjacent difference closest to the cluster average value corresponds toThe value of (2) is recorded as a first threshold.
6. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the obtaining the density data distance corresponding to the historical data according to the data distance of the historical data and the first threshold value comprises the following specific steps:
and respectively taking each historical data in the historical data set as the historical data to be analyzed, selecting a first threshold historical data with the smallest data distance with the historical data to be analyzed in the historical data set, and recording the sum of the data distances of the selected first threshold historical data value and the historical data to be analyzed as the density data distance corresponding to the historical data to be analyzed.
7. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the determining the density distance between two historical data corresponding to the same data category in two simultaneous historical data sets according to the density data distance corresponding to the historical data contained in the historical data sets comprises the following specific steps:
recording absolute values of differences of density data distances corresponding to two historical data corresponding to the same data type in the two simultaneous historical data sets as first absolute values;
the first absolute value is noted as the density distance.
8. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the specific method for acquiring the evaluation distance is as follows:
in the method, in the process of the invention,for the simultaneous history data set->And time history data set->The evaluation distance between the two; />For the simultaneous history data set->And time history data set->Corresponding to the same data category->Density distance between two historic data of (1), wherein +.>；/>The number of different data types contained in the historical data set at the same moment; />Data category->And data category->And (5) adjusting the coefficient of the data category association degree.
9. The historical data analysis-based heating, ventilation and energy saving ladder optimizing control method according to claim 1, wherein each equipment of the heating, ventilation and air conditioning comprises a field control cabinet, a cooling tower, a cooling pump, a water chilling unit, a freezing pump and a plate heat exchanger.
10. The heating ventilation energy-saving ladder optimizing control method based on historical data analysis according to claim 1, wherein the method for obtaining the optimal control parameters of each device of the heating ventilation air conditioner by using an AI algorithm system according to the evaluation distance between the historical data sets at the same time to realize the ladder optimizing control comprises the following specific steps:
the step optimizing control process of the heating ventilation air conditioner comprises 5 stages, namely L1-L5, the L1-L5 process is gradually optimized and analyzed through reinforcement learning by using an AI algorithm system, and after analysis results are implemented, feedback results are fed back to the AI algorithm system, so that continuous updating iteration and control optimization are realized;
the L1 stage is used for carrying out monomer optimization on all equipment of the heating ventilation air conditioner, including but not limited to cooling tower power optimization of a cold source system, water pump frequency optimization, chiller power optimization, control logic of a precise air conditioner and optimization of fan rotating speed;
the L2 stage is used for carrying out combined optimization on all the monomer optimizing devices of the L1, and specifically comprises group control optimizing under a single refrigeration unit and under a combined and coordinated working condition of all the devices and group control optimizing of a precise air conditioning unit;
the L3 stage is used for performing joint debugging on the single refrigeration unit in the L2 stage, and optimizing among the multiple refrigeration units, including but not limited to optimizing the starting quantity of the units of the cold source control system, optimizing the unit load distribution and controlling and optimizing the round-robin work among the multiple refrigeration units;
the L4 stage is used for optimizing a plurality of sets of refrigeration units in the L3 and optimizing the tail end precise air conditioner in a linkage manner, and determining optimal chilled water supply water temperature and tail end air quantity parameters;
and the L5 stage is used for carrying out full-system optimization on the IT equipment air-conditioning system, the movable ring system, the refrigeration system and the energy storage system by carrying out full-equipment linkage on the IT equipment air-conditioning system, the movable ring system, the refrigeration system and the energy storage system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311204306.4A CN116955963B (en) | 2023-09-19 | 2023-09-19 | Heating ventilation energy-saving ladder optimizing control method based on historical data analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311204306.4A CN116955963B (en) | 2023-09-19 | 2023-09-19 | Heating ventilation energy-saving ladder optimizing control method based on historical data analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116955963A true CN116955963A (en) | 2023-10-27 |
| CN116955963B CN116955963B (en) | 2023-12-08 |
Family
ID=88453243
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311204306.4A Active CN116955963B (en) | 2023-09-19 | 2023-09-19 | Heating ventilation energy-saving ladder optimizing control method based on historical data analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116955963B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117272209A (en) * | 2023-11-20 | 2023-12-22 | 江苏新希望生态科技有限公司 | Bud seedling vegetable growth data acquisition method and system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130031093A1 (en) * | 2011-07-25 | 2013-01-31 | Sony Computer Entertainment Inc. | Information processing system, information processing method, program, and non-transitory information storage medium |
| KR101396394B1 (en) * | 2013-03-20 | 2014-05-19 | 주식회사 스마티랩 | Methods to autonomously optimize performance using clustering in mobile cloud environment |
| CN109271424A (en) * | 2018-09-29 | 2019-01-25 | 海南大学 | A kind of parameter adaptive clustering method based on density |
| CN113408659A (en) * | 2021-07-15 | 2021-09-17 | 重庆大学 | Building energy consumption integrated analysis method based on data mining |
-
2023
- 2023-09-19 CN CN202311204306.4A patent/CN116955963B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130031093A1 (en) * | 2011-07-25 | 2013-01-31 | Sony Computer Entertainment Inc. | Information processing system, information processing method, program, and non-transitory information storage medium |
| KR101396394B1 (en) * | 2013-03-20 | 2014-05-19 | 주식회사 스마티랩 | Methods to autonomously optimize performance using clustering in mobile cloud environment |
| CN109271424A (en) * | 2018-09-29 | 2019-01-25 | 海南大学 | A kind of parameter adaptive clustering method based on density |
| CN113408659A (en) * | 2021-07-15 | 2021-09-17 | 重庆大学 | Building energy consumption integrated analysis method based on data mining |
Non-Patent Citations (2)
| Title |
|---|
| YUQING YANG等: "Density clustering with divergence distance and automatic center selection", 《INFORMATION SCIENCE》, vol. 596 * |
| 陈晋音;何辉豪;: "基于密度的聚类中心自动确定的混合属性数据聚类算法研究", 自动化学报, no. 10 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117272209A (en) * | 2023-11-20 | 2023-12-22 | 江苏新希望生态科技有限公司 | Bud seedling vegetable growth data acquisition method and system |
| CN117272209B (en) * | 2023-11-20 | 2024-02-02 | 江苏新希望生态科技有限公司 | Bud seedling vegetable growth data acquisition method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116955963B (en) | 2023-12-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106920006B (en) | Subway station air conditioning system energy consumption prediction method based on ISOA-LSSVM | |
| CN104566868B (en) | A kind of central air conditioning system and its control method | |
| WO2023093820A1 (en) | Device control optimization method, display platform, cloud server, and storage medium | |
| CN104534627B (en) | Central air conditioning cooling water system comprehensive energy efficiency control method | |
| CN110726218B (en) | Air conditioner, control method and device thereof, storage medium and processor | |
| CN116955963B (en) | Heating ventilation energy-saving ladder optimizing control method based on historical data analysis | |
| CN104019526B (en) | Improve PSO algorithm Fuzzy Adaptive PID temperature and humidity control system and method | |
| CN111780332B (en) | Household metering method and device for central air conditioner | |
| CN107044710A (en) | Energy-saving control method for central air conditioner and system based on joint intelligent algorithm | |
| CN110567101A (en) | Water chiller high-energy-efficiency control method based on support vector machine model | |
| CN113739365A (en) | Central air-conditioning cold station group control energy-saving control method, device, equipment and storage medium | |
| CN1482409A (en) | Central air-conditioning energy conserving fuzzy controlling method and fuzzy controller thereof | |
| CN108168030A (en) | A kind of intelligent control method based on refrigeration performance curve | |
| CN113268913B (en) | Intelligent building air conditioner cooling machine system operation optimization method based on PSO-ELM algorithm | |
| CN111830350B (en) | Energy consumption metering method and device and electric appliance | |
| CN106642553A (en) | Air conditioner control method | |
| CN113778215A (en) | Method for realizing data center PUE prediction and consumption reduction strategy based on big data | |
| CN112113314A (en) | Real-time temperature data acquisition system and temperature adjusting method based on learning model | |
| CN112292013A (en) | Micro-module data center heat source prejudgment and cold source regulation and control method based on time sequence | |
| CN114970358A (en) | Data center energy efficiency optimization method and system based on reinforcement learning | |
| CN110726216B (en) | Air conditioner, control method, device and system thereof, storage medium and processor | |
| CN113188585B (en) | Freezing station sensor fault diagnosis method based on few redundant sensors | |
| CN113959071B (en) | Centralized water chilling unit air conditioning system operation control optimization method based on machine learning assistance | |
| CN105240993A (en) | Fine energy-saving control system of central air conditioner and achieving method of fine energy-saving control system | |
| CN204853838U (en) | Central air conditioning control system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |