CN113048549A - Heating circulating pump adjusting method based on artificial intelligence - Google Patents
Heating circulating pump adjusting method based on artificial intelligence Download PDFInfo
- Publication number
- CN113048549A CN113048549A CN202110285916.6A CN202110285916A CN113048549A CN 113048549 A CN113048549 A CN 113048549A CN 202110285916 A CN202110285916 A CN 202110285916A CN 113048549 A CN113048549 A CN 113048549A
- Authority
- CN
- China
- Prior art keywords
- target
- network
- circulating pump
- pressure difference
- model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 23
- 238000010438 heat treatment Methods 0.000 title claims abstract description 22
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 13
- 238000004140 cleaning Methods 0.000 claims abstract description 6
- 230000002787 reinforcement Effects 0.000 claims abstract description 5
- 230000002159 abnormal effect Effects 0.000 claims description 3
- 230000000737 periodic effect Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims 1
- 238000012549 training Methods 0.000 abstract description 2
- 230000032683 aging Effects 0.000 description 2
- 230000003750 conditioning effect Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24D—DOMESTIC- OR SPACE-HEATING SYSTEMS, e.g. CENTRAL HEATING SYSTEMS; DOMESTIC HOT-WATER SUPPLY SYSTEMS; ELEMENTS OR COMPONENTS THEREFOR
- F24D19/00—Details
- F24D19/10—Arrangement or mounting of control or safety devices
- F24D19/1006—Arrangement or mounting of control or safety devices for water heating systems
- F24D19/1009—Arrangement or mounting of control or safety devices for water heating systems for central heating
- F24D19/1012—Arrangement or mounting of control or safety devices for water heating systems for central heating by regulating the speed of a pump
Abstract
The invention discloses a heating circulating pump adjusting method based on artificial intelligence, which comprises the specific steps of obtaining target pressure difference, target two-network flow and real-time online data of a heating unit, and cleaning and normalizing the real-time data; the method comprises the steps of learning a prediction model of a reinforcement learning strategy model on line, and updating the frequency of a circulating pump of the model on line according to a difference value between a target pressure difference and an actual pressure difference or a difference value between a target second-network flow and an actual second-network flow; and determining the minimum difference value, inputting the minimum difference value into a prediction model to obtain the frequency of the circulating pump, and further adjusting the circulating pump. The invention adopts reinforcement learning, and the relationship between the frequency of the circulating pump and the pressure difference and the flow of the second network maintains a stable value of the pressure difference or the flow of the second network through real-time online learning, thereby having strong generalization capability. And training the model by using historical data in the database, learning a strategy which can obtain a good result, inputting the real-time state into the model, and outputting an optimal action to adjust the frequency of the circulating pump.
Description
Technical Field
The invention relates to the technical field of resident heating, in particular to a heating circulating pump adjusting method based on artificial intelligence.
Background
The traditional method for adjusting and controlling the frequency of the circulating pump basically adopts an automatic controller plc built-in pid algorithm, and sets three parameter values of proportion, integral and differential by setting a sampling period, so that the pressure difference or the flow of a secondary pipe network is maintained at a constant value. The method needs a great deal of time for an engineer with strong professional ability to set the parameters, the parameter setting cannot cause pressure difference or great fluctuation of flow, and the generalization ability of the parameter values is poor.
The prior art is that the pid control algorithm that uses, to engineer's requirement than higher, the model of the circulating pump that different heat exchange stations used is different, and when voltage is unstable, the circulating pump is ageing, and secondary pipe network is ageing, all can cause very big influence to pid parameter setting, needs the engineer to go to update the parameter of pid regularly, and when having thousands of units to a heating power company, this work load will be huge.
Disclosure of Invention
The invention aims to overcome the defects in the prior art, and in order to realize the aim, the heating circulating pump adjusting method based on artificial intelligence is adopted to solve the problems in the background technology.
A heating circulating pump adjusting method based on artificial intelligence comprises the following specific steps:
acquiring target pressure difference, target two-network flow and real-time online data of a heating unit, and cleaning and normalizing the real-time data;
the method comprises the steps of learning a prediction model of a reinforcement learning strategy model on line, and updating the frequency of a circulating pump of the model on line according to a difference value between a target pressure difference and an actual pressure difference or a difference value between a target second-network flow and an actual second-network flow;
and determining the minimum difference value, inputting the minimum difference value into a prediction model to obtain the frequency of the circulating pump, and further adjusting the circulating pump.
As a further aspect of the invention: the specific steps of updating the frequency of the circulating pump of the model on line according to the difference value between the target pressure difference and the actual pressure difference or the difference value between the target second-network flow and the actual second-network flow comprise:
the formula for determining the target differential pressure according to the differential pressure at the current moment is as follows:
or the formula for determining the target second-network flow by the current second-network flow is as follows:
wherein, PtFor the pressure difference at the present moment, PtargetIs a target pressure difference, PΔtFor the next moment of pressure difference, TtFor the current time two-network traffic, TtargetFor target two-network traffic, TΔtThe traffic of the second network at the next time.
As a further aspect of the invention: the specific steps of cleaning and normalizing the real-time data comprise:
removing abnormal invalid data from the acquired real-time data;
the data is then normalized and mapped to interval (0, 1).
As a further aspect of the invention: the specific steps of inputting the prediction model to obtain the circulating pump frequency comprise:
controlling the frequency of the circulating pump according to the real-time state data through a periodic operation model;
the predicted pressure difference between the supply pressure and the return pressure of the two networks is consistent with the target pressure difference; or
And (4) if the flow of the second network is consistent with the target flow, acquiring the frequency of the circulating pump and further adjusting.
As a further aspect of the invention: the policy network update formula is as follows:
where s is a state characteristic of the environment, stIs the state at timestamp t; a is the action taken by the agent, atRepresents an action at a time stamp t; pi (a | s) is a decision model of the agent, s is a state of receiving input, p (a | s) represents probability distribution of executing action after decision is given, and sigma is satisfieda∈APi (a | s), r (a | s) is a feedback signal given after the environment has received action a in state s, and the reward obtained at time stamp t is denoted as rt，
The value is output for the value network.
As a further aspect of the invention: the value network parameter update formula is:
wherein the content of the first and second substances,
a distance measure representing a value network output value, dist (a, b) being a and b, in Euclidean distance;
is a target value of; gamma is the learning rate of the value network;
and collecting real-time operation data of the circulating pump, and predicting the output flow of the circulating pump or the hydraulic pressure difference target behind the pump in front of the pump by combining a model.
Compared with the prior art, the invention has the following technical effects:
by adopting the technical scheme, the real-time data is preprocessed through the target pressure difference, the target two-network flow and the real-time online data of the heating unit. Along with the change of the environment, the parameters need to be updated in time. Meanwhile, the strategy model can be automatically updated regularly, the generalization capability is strong, and the method can adapt to the change of the environment. Without excessive human intervention.
Drawings
The following detailed description of embodiments of the invention refers to the accompanying drawings in which:
FIG. 1 is a schematic step diagram of a heating circulation pump conditioning method according to some embodiments disclosed herein;
FIG. 2 is a block flow diagram of a heating circulation pump conditioning method of some embodiments disclosed herein.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Referring to fig. 1 and fig. 2, in an embodiment of the present invention, a heating circulation pump adjusting method based on artificial intelligence includes:
s1, acquiring a target pressure difference, a target two-network flow and real-time online data of the heating unit, and cleaning and normalizing the real-time data;
specifically, extracting real-time data required by the model, removing abnormal invalid data from the acquired real-time data according to the model requirement, sequencing according to the acquisition time, and constructing data with the time interval of 1 second;
and then all data are normalized by max-min and mapped to the interval (0, 1). According to the data characteristics, sample data is divided into a training set, a testing set and a verification set according to 70%, 20% and 10%.
S2, online learning a prediction model of the reinforcement learning strategy model, and updating the frequency of a circulating pump of the model online according to the difference value between the target pressure difference and the actual pressure difference or the difference value between the target second-network flow and the actual second-network flow;
the method comprises the following specific steps:
the formula for determining the target differential pressure according to the differential pressure at the current moment is as follows:
or the formula for determining the target second-network flow by the current second-network flow is as follows:
wherein, PtFor the pressure difference at the present moment, PtargetIs a target pressure difference, PΔtFor the next moment of pressure difference, TtFor the current time two-network traffic, TtargetFor target two-network traffic, TΔtThe traffic of the second network at the next time.
And S3, determining that the difference is minimum, inputting the difference into the prediction model to obtain the frequency of the circulating pump, and further adjusting the circulating pump.
The method comprises the following specific steps:
controlling the frequency of the circulating pump according to the acquired real-time state data of the circulating pump through a periodic operation model;
the predicted pressure difference between the supply pressure and the return pressure of the two networks is consistent with the target pressure difference; or the flow of the second network is consistent with the target flow, the frequency of the circulating pump is obtained and further adjusted, and the obtained frequency of the circulating pump is the frequency of the circulating pump obtained by the model reverse-deducing.
In some specific embodiments, the policy network update formula of the model is:
where s is a state characteristic of the environment, stIs the state at timestamp t; a is the action taken by the agent, atRepresents an action at a time stamp t; pi (a | s) is a decision model of the agent, s is a state of receiving input, p (a | s) represents probability distribution of executing action after decision is given, and sigma is satisfieda∈APi (a | s), r (a | s) is a feedback signal given after the environment has received action a in state s, and the reward obtained at time stamp t is denoted as rt，
The value is output for the value network.
In some specific embodiments, the value network parameter update formula is:
wherein the content of the first and second substances,
a distance measure representing a value network output value, dist (a, b) being a and b, in Euclidean distance;
is a target value of; gamma is the learning rate of the value network;
and collecting real-time operation data of the circulating pump, and predicting the output flow of the circulating pump or the hydraulic pressure difference target behind the pump in front of the pump by combining a model.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents, which should be construed as being within the scope of the invention.
Claims (6)
1. A heating circulating pump adjusting method based on artificial intelligence is characterized by comprising the following steps:
acquiring target pressure difference, target two-network flow and real-time online data of a heating unit, and cleaning and normalizing the real-time data;
the method comprises the steps of learning a prediction model of a reinforcement learning strategy model on line, and updating the frequency of a circulating pump of the model on line according to a difference value between a target pressure difference and an actual pressure difference or a difference value between a target second-network flow and an actual second-network flow;
and determining the minimum difference value, inputting the minimum difference value into a prediction model to obtain the frequency of the circulating pump, and further adjusting the circulating pump.
2. The heating circulation pump adjusting method based on artificial intelligence of claim 1, wherein the specific step of updating the frequency of the circulation pump of the model on line according to the difference between the target pressure difference and the actual pressure difference or the difference between the target two-network flow and the actual two-network flow comprises:
the formula for determining the target differential pressure according to the differential pressure at the current moment is as follows:
or the formula for determining the target second-network flow by the current second-network flow is as follows:
wherein, PtFor the pressure difference at the present moment, PtargetIs a target pressure difference, PΔtFor the next moment of pressure difference, TtFor the current time two-network traffic, TtargetFor target two-network traffic, TΔtThe traffic of the second network at the next time.
3. The artificial intelligence-based heating circulation pump adjusting method according to claim 1, wherein the specific steps of performing cleaning normalization on the real-time data include:
removing abnormal invalid data from the acquired real-time data;
the data is then normalized and mapped to interval (0, 1).
4. The artificial intelligence based heating circulation pump adjusting method according to claim 3, wherein the specific step of inputting the predictive model to obtain the circulation pump frequency comprises:
controlling the frequency of the circulating pump according to the real-time state data through a periodic operation model;
the predicted pressure difference between the supply pressure and the return pressure of the two networks is consistent with the target pressure difference; or
And (4) if the flow of the second network is consistent with the target flow, acquiring the frequency of the circulating pump and further adjusting.
5. The artificial intelligence based heating circulation pump adjusting method according to claim 4, wherein the strategy network update formula is as follows:
where s is a state characteristic of the environment, stIs the state at timestamp t; a is the action taken by the agent, atRepresents an action at a time stamp t; pi (a | s) is a decision model of the agent, s is a state of receiving input, p (a | s) represents probability distribution of executing action after decision is given, and sigma is satisfieda∈APi (a | s), r (a | s) is a feedback signal given after the environment has received action a in state s, and the reward obtained at time stamp t is denoted as rt，
The value is output for the value network.
6. The artificial intelligence based heating circulation pump adjusting method according to claim 5, wherein the value network parameter update formula is:
wherein the content of the first and second substances,
a distance measure representing a value network output value, dist (a, b) being a and b, in Euclidean distance;
is a target value of; gamma is the learning rate of the value network;
and collecting real-time operation data of the circulating pump, and predicting the output flow of the circulating pump or the hydraulic pressure difference target behind the pump in front of the pump by combining a model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110285916.6A CN113048549A (en) | 2021-03-17 | 2021-03-17 | Heating circulating pump adjusting method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110285916.6A CN113048549A (en) | 2021-03-17 | 2021-03-17 | Heating circulating pump adjusting method based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113048549A true CN113048549A (en) | 2021-06-29 |
Family
ID=76513105
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110285916.6A Pending CN113048549A (en) | 2021-03-17 | 2021-03-17 | Heating circulating pump adjusting method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113048549A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008039200A (en) * | 2006-08-01 | 2008-02-21 | Sanyo Electric Co Ltd | Refrigerating device |
| KR101466652B1 (en) * | 2013-12-23 | 2014-11-28 | 한국건설기술연구원 | Central cooling and heating system and its control method using differential pressure and temperature for multi-family houses |
| TW201503623A (en) * | 2013-03-15 | 2015-01-16 | Interdigital Patent Holdings | Methods and procedures for non-linear precoding based multiuser multiple input multiple output |
| CN109185972A (en) * | 2018-09-11 | 2019-01-11 | 哈尔滨顺易天翔热力技术开发有限公司 | Thermal substation circulating pump optimal regulation method |
| CN109245112A (en) * | 2018-10-18 | 2019-01-18 | 南京丰道电力科技有限公司 | Determination method and device for the reactive compensation power that photovoltaic plant voltage is adjusted |
| CN110274302A (en) * | 2019-07-10 | 2019-09-24 | 合肥瑞纳节能工程有限公司 | A kind of wisdomization heat exchange station control system |
| CN111076376A (en) * | 2019-12-18 | 2020-04-28 | 西安建筑科技大学 | Method and system for predicting cold load demand and distributing ice storage air conditioner load |
| CN111561732A (en) * | 2020-05-18 | 2020-08-21 | 瑞纳智能设备股份有限公司 | Heat exchange station heat supply adjusting method and system based on artificial intelligence |
| CN111678195A (en) * | 2020-04-30 | 2020-09-18 | 太原大四方节能环保股份有限公司 | Regulation and control method of quality branch pump heat supply unit |
| CN112254237A (en) * | 2020-09-21 | 2021-01-22 | 无锡混沌能源技术有限公司 | Air conditioner circulating water system pressure difference control system |
-
2021
- 2021-03-17 CN CN202110285916.6A patent/CN113048549A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2008039200A (en) * | 2006-08-01 | 2008-02-21 | Sanyo Electric Co Ltd | Refrigerating device |
| TW201503623A (en) * | 2013-03-15 | 2015-01-16 | Interdigital Patent Holdings | Methods and procedures for non-linear precoding based multiuser multiple input multiple output |
| KR101466652B1 (en) * | 2013-12-23 | 2014-11-28 | 한국건설기술연구원 | Central cooling and heating system and its control method using differential pressure and temperature for multi-family houses |
| CN109185972A (en) * | 2018-09-11 | 2019-01-11 | 哈尔滨顺易天翔热力技术开发有限公司 | Thermal substation circulating pump optimal regulation method |
| CN109245112A (en) * | 2018-10-18 | 2019-01-18 | 南京丰道电力科技有限公司 | Determination method and device for the reactive compensation power that photovoltaic plant voltage is adjusted |
| CN110274302A (en) * | 2019-07-10 | 2019-09-24 | 合肥瑞纳节能工程有限公司 | A kind of wisdomization heat exchange station control system |
| CN111076376A (en) * | 2019-12-18 | 2020-04-28 | 西安建筑科技大学 | Method and system for predicting cold load demand and distributing ice storage air conditioner load |
| CN111678195A (en) * | 2020-04-30 | 2020-09-18 | 太原大四方节能环保股份有限公司 | Regulation and control method of quality branch pump heat supply unit |
| CN111561732A (en) * | 2020-05-18 | 2020-08-21 | 瑞纳智能设备股份有限公司 | Heat exchange station heat supply adjusting method and system based on artificial intelligence |
| CN112254237A (en) * | 2020-09-21 | 2021-01-22 | 无锡混沌能源技术有限公司 | Air conditioner circulating water system pressure difference control system |
Non-Patent Citations (1)
| Title |
|---|
| 邹锋: ""基于生成对抗网络的强化学习方法及其在建筑节能方面的应用"", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, no. 01, pages 7 - 34 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20240000046A1 (en) | Predictive control system and regulatory method for temperature of livestock house | |
| CN109270842B (en) | Bayesian network-based regional heat supply model prediction control system and method | |
| WO2021196997A1 (en) | Method for predicting and controlling water level of series of water conveyance canals on basis of fuzzy neural network | |
| CN111580382B (en) | Unit-level heat supply adjusting method and system based on artificial intelligence | |
| CN111561732B (en) | Heat exchange station heat supply adjusting method and system based on artificial intelligence | |
| CN115962551B (en) | Intelligent air conditioner control system and method for building automatic control | |
| CN111811016B (en) | Method and system for adjusting hydraulic balance among units of heating unit | |
| CN112413831A (en) | Energy-saving control system and method for central air conditioner | |
| CN106468467B (en) | Air conditioner cold load real-time prediction algorithm applied to embedded control system | |
| CN111520812B (en) | Method and system for estimating room temperature of heat supply residents | |
| CN117093823B (en) | Factory energy consumption management method based on data analysis | |
| CN113110056B (en) | Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence | |
| CN114781249A (en) | High-density clarification tank dosage prediction and control method based on multidimensional scoring model | |
| CN114484556A (en) | Water supply temperature regulation and control method based on target energy consumption management and control | |
| CN113888202A (en) | Training method and application method of electricity price prediction model | |
| CN113048549A (en) | Heating circulating pump adjusting method based on artificial intelligence | |
| CN111062576A (en) | DMA partition based pressure management system with least disadvantages | |
| CN110633844B (en) | Building energy system simulation prediction method based on EMD and ANN and application | |
| CN114282445A (en) | Circulating water system operation optimization method and related equipment | |
| CN114110716A (en) | Operation energy efficiency monitoring control system and method for thermoelectric unit, computer equipment and storage medium | |
| CN116599061B (en) | Power grid operation control method based on reinforcement learning | |
| CN113065772B (en) | Method for constructing heating system regulation and control target model based on k-means clustering algorithm | |
| CN117273402B (en) | Energy-saving management system and method for glass deep processing production line based on Internet of Things technology | |
| CN117406685B (en) | Intelligent control optimizing management system of building equipment suitable for green low-carbon building | |
| CN114543273B (en) | Self-adaptive deep learning optimization energy-saving control algorithm for central air conditioner cooling 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 |