CN113110056B - Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence - Google Patents
Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence Download PDFInfo
- Publication number
- CN113110056B CN113110056B CN202110426947.9A CN202110426947A CN113110056B CN 113110056 B CN113110056 B CN 113110056B CN 202110426947 A CN202110426947 A CN 202110426947A CN 113110056 B CN113110056 B CN 113110056B
- Authority
- CN
- China
- Prior art keywords
- temperature
- room temperature
- water supply
- controllable quantity
- artificial intelligence
- 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention relates to the technical field of artificial intelligence heat supply, in particular to a heat supply intelligent decision method based on artificial intelligence, which comprises the following steps: acquiring a data set of equivalent room temperature tn of at least one target space and a corresponding relation between the indoor temperature tn and a controllable quantity tc of a heating station at a heat supply position; and obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter. The scheme can adapt to the characteristic that the input quantity of the actual system is chaotic or probabilistic, output the optimal value distribution, adapt to the characteristic that the input quantity of the actual system is chaotic or probabilistic, realize the intelligent regulation and control of the water supply temperature and have high precision.
Description
Technical Field
The invention relates to the technical field of artificial intelligence heat supply, in particular to a heat supply intelligent decision-making method and an intelligent decision-making machine based on artificial intelligence.
Background
Along with the continuous expansion of the urban scale, the coverage area and the heat supply area of an urban centralized heat supply network are larger and larger, and the number of heat stations is more and more; it is difficult to realize in time that the heat supply as required causes very big wasting of resources, and user's comfort level is also difficult to guarantee. In the traditional heat supply, two methods of calculating by formula or determining heat load of the heat station under different outdoor temperatures in different areas according to experience are mainly adopted to regulate heat supply of the heat station, but the calculation based on the complex formula needs to be performed under the ideal condition, and the actual condition is greatly different from the ideal condition; meanwhile, the heat load of the heating power station in different areas at different outdoor temperatures is determined according to experience, and the dependence on the experience is large; the two methods are low in efficiency and poor in applicability, and have large errors.
For example, a central heating secondary network operation adjusting method with application number of CN201310584999.4 relates to the technical field of heating adjustment. And a method of combining a three-layer forward neural network and PID is adopted to realize automatic control of the return water temperature of the secondary network. The method comprises the steps of establishing a set value prediction model of the central heating secondary network temperature control system by using a RBF neural network, predicting the secondary network return water temperature in real time by using five data of primary network supply water temperature, primary network supply water flow, outdoor temperature, secondary network supply water temperature and secondary network supply water flow collected on site through the trained RBF prediction model, and taking the value as the set value of the central heating secondary network temperature control system. The scheme is closed-loop control, is single in control, cannot control the temperatures of a plurality of rooms with different requirements simultaneously, and cannot predict the equivalent room temperatures of different rooms.
The application number is CN202010419825.2, a heating power station heating regulation method and system based on artificial intelligence, firstly, defining relevant parameters and combining with an intelligent agent to generate operation data in the actual heating process and preprocessing the operation data; then, performing model training on the processed operation data through an artificial intelligence algorithm to obtain a prediction model; then collecting real-time operation data and predicting target values of secondary flow and secondary temperature supply by combining the model; and finally, controlling the frequency of the circulating pump and the opening of the one-network valve to adjust the heat load of the heating power station according to the predicted target value, thereby realizing the dynamic adjustment of the indoor uniform temperature. The control input variable of the scheme is constant and single, and the optimal solution between the demands of multiple room temperatures and the total supply quantity of the power supply station cannot be searched simultaneously, so that energy waste or poor user experience is caused.
Disclosure of Invention
The purpose of the invention is: the heat supply intelligent decision method and the intelligent decision machine based on artificial intelligence solve the technical problem that an existing model cannot simultaneously seek the optimal solution between the demands of multiple room temperatures and the total supply quantity of a power supply station.
The technical scheme of the invention is as follows: a heat supply intelligent decision-making method based on artificial intelligence comprises the following steps:
acquiring a data set of equivalent room temperature tn of at least one target space and a corresponding relation between the room temperature tn and a controllable quantity tc of a heat station at a heat supply position.
And obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter.
Optionally, a stochastic model is used to select the indoor temperature of at least one space.
Optionally, the heat station controllable quantity tc is a single physical controllable quantity or a plurality of controllable quantity sets.
Optionally, the controllable quantity tc of the thermal station is one or more of a primary side water supply temperature, a return water temperature, a primary side water supply pressure, a primary side electric regulating valve position, a water supply and return average temperature, a primary side flow, a primary side distributed pump frequency, a secondary side water supply temperature, a secondary side water supply pressure, a secondary side flow, a secondary side pump frequency, a return water pressure or heat at a heat supply position.
Optionally, when the water supply temperature is decided in each control period, the prediction model P is used as an object, the target room temperature ts is used as a target, tc is optimized, the optimal tc is found, and tc is used as a controllable quantity target to be output.
Optionally, the input parameter is a single value, a distribution function or a data table.
Optionally, the output parameter is a single value, an optimal value distribution function, or a data table.
Optionally, with the target room temperature as a target, obtaining an optimal controllable amount tc output parameter of the thermal station according to the prediction model P to control heat supply at the heat supply position.
The other technical scheme of the invention is as follows: an intelligent decision machine for a heating intelligent decision method based on artificial intelligence, comprising: the system comprises a CPU and one or more information collectors, wherein the information collectors are used for acquiring equivalent room temperature tn of a target space and controllable amount tc of a heating station at a heat supply position corresponding to the equivalent room temperature tn; the information collector may be a sensor directly deployed at a user side, or may be an interface for receiving user side information through a network.
The CPU is used for obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter.
Has the advantages that: the invention provides a heat supply intelligent decision method based on artificial intelligence, which comprises the following steps: acquiring a data set of equivalent room temperature tn of at least one target space and a corresponding relation between the indoor temperature tn and a controllable quantity tc of a heating station at a heat supply position; and obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter. The scheme can adapt to the characteristic that the input quantity of the actual system is chaotic or probabilistic, output the optimal value distribution, adapt to the characteristic that the input quantity of the actual system is chaotic or probabilistic, realize the intelligent regulation and control of the water supply temperature and have high precision.
Drawings
FIG. 1 is a schematic flow chart of an artificial intelligence-based intelligent heat supply decision method according to the present invention;
fig. 2 is a diagram of the use effect of the heat supply intelligent decision machine based on artificial intelligence.
Detailed Description
Embodiment 1, as shown in fig. 1, a heat supply intelligent decision method based on artificial intelligence includes: acquiring a data set of equivalent room temperature tn of at least one target space and a corresponding relation between the room temperature tn and a controllable quantity tc of a heat station at a heat supply position.
And obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter.
The automatic control parameters at the heat supply place and the heating power station mainly comprise control variables such as water supply temperature, return water temperature, supply and return water (weighted) average temperature, flow, heat and the like, and the control variables are controllable quantity tc of the heating power station. The target variable is mainly the room temperature, namely the equivalent room temperature tn, namely at least one room temperature, and the control variable of the heat supply part is autonomously learned and adjusted through an algorithm and a strategy link, so that the final result shows that the room temperature reaches the expectation of a permanent user. The algorithm and strategy links adopt machine learning and artificial intelligence technology.
Specifically, in a sampling control period, the equivalent room temperature tn of one or more target spaces and the controllable amount tc of the thermal station at the heat supply position in the sampling control period are obtained, so that the corresponding relationship between the equivalent room temperature tn and the controllable amount tc of the thermal station can be obtained. So that the corresponding relation between a plurality of target spaces and the controllable quantity tc of the thermal station in different time periods can be obtained. In this way a data set is obtained. And then, obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter. Namely, the data sets are trained by using a machine learning technology, namely an artificial intelligence algorithm, so that the relation between the equivalent room temperature and the controllable quantity and the outdoor temperature of the thermal power station is obtained, and a prediction model P is established.
Optionally, the controllable amount tc of the thermal station is a single physical controllable amount or a plurality of controllable amount sets. It should be noted that the controllable quantity tc of the thermal power station includes at least one controllable variable, and the specific controllable variable can be selected according to actual needs.
Optionally, the controllable amount tc of the thermal station is one or more of a primary side water supply temperature, a return water temperature, a primary side water supply pressure, a primary side electric regulating valve position, a water supply and return average temperature, a primary side flow, a primary side distributed pump frequency, a secondary side water supply temperature, a secondary side water supply pressure, a secondary side flow, a secondary side pump frequency, a return water pressure, or heat of a heat supply place. The thermal station controllable quantity tc may correspond to a single physically controllable quantity, such as the water supply temperature tg, or may be a plurality of controllable quantities mapped or derived by a logic or function.
Optionally, a stochastic model is used to select the indoor temperature of at least one space. First, a certain mathematical statistical model is used to select the room temperature, such as a random model. The calculation for the selected room temperature is then based on the equivalent room temperature tn for one or more rooms. Such as average room temperature. Other statistical approaches are possible, such as median room temperature, distribution density weighted average room temperature, average room temperature inferred assuming normal distribution, room temperature weighted for regulatory requirements, etc.
And collecting a data set of a certain amount of equivalent room temperature tn and the controllable amount tc of the thermal station in a corresponding relationship. The controllable quantity tn may correspond to a single physically controllable quantity, such as the supply water temperature tg, or may be a plurality of controllable quantities mapped or derived by a logic or function. The physical controllable quantity tm may be one or more derived quantities of the supply water temperature tg, the return water temperature th, the supply and return water (weighted) average temperature tp, the flow Q, and the heat Q.
And processing the data sets by using a machine learning technology, namely an artificial intelligence algorithm, obtaining the relation between the equivalent room temperature and the controllable quantity and the outdoor temperature, and establishing a prediction model P. I.e. this model P has the ability to input different thermal station controlled quantities tc, predicting the equivalent room temperature tn for different rooms.
Optionally, a data set of an equivalent room temperature tn of at least one target space and a correspondence between the room temperature tn and a controllable amount tc of the thermal station at the heat supply site is obtained.
And (3) obtaining a prediction model P by training by adopting an enhanced learning method (RL) and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter.
The prediction model P is trained by taking the prediction model P as an Environment (Environment) of the reinforcement learning method, taking the controllable quantity tc of the thermal station as an Action (Action) of the reinforcement learning method and taking a function value (Reward) of which the equivalent room temperature tn is in inverse proportion to the absolute value of the difference between the target room temperature ts as a return.
After the optimal controllable quantity tc is determined, the functional relation tm ═ fm (tc) is substituted, and the physical controllable quantity tm can be derived from tc. The physical controllable quantity tm may be one or more derived quantities of the supply water temperature tg, the return water temperature th, the supply and return water (weighted) average temperature tp, the flow Q, and the heat Q.
In a specific implementation scenario, a deep neural network method in an artificial intelligence algorithm is adopted, the controllable quantity tc of the thermal power station is used as the input of the neural network, the indoor temperature tn is used as the output, and the prediction model P is obtained through training. And when the water supply temperature is decided in each control period, optimizing the controllable quantity tc of the thermal power station by taking the prediction model P as an object and the target room temperature ts as a target to find out the optimal tc. And outputting tc as a controllable quantity target.
Optionally, the input parameter is a single value, a distribution function, or a data table. The input parameter of the present control method may be a distribution or a distribution probability. In the traditional input parameters, the backwater temperature and the outdoor temperature are both determined values, and in the algorithm, the quantities can be distribution, and the expression form can be a distribution function, a data table and the like. The characteristic makes the algorithm adapt to the characteristic that the input quantity of the actual system is chaotic or probabilistic. In practical applications, such quantities are common in practice, and for example, the return water temperature, the outdoor temperature, and the like are actually a variable quantity.
Optionally, the output parameter is a single value, an optimal value distribution function, or a data table. Unlike the conventional method, the output decision quantity of the algorithm may not be a single value but may be a distribution of optimal values. For example, when the algorithm is used for deciding the temperature of the secondary network water supply, the given decision output is not a single value, but the distribution of different optimal secondary network water supply temperatures. The output capability of the algorithm can solve the problem that global optimization becomes possible when the control link is linked with other upper-layer optimization decision links.
In an optional scheme, the target room temperature is used as a target, and an optimal controllable amount tc output parameter of the heating station is obtained according to the prediction model P so as to control heat supply at a heating position. The prediction model P can be applied in any occasions, and the optimal controllable quantity tc output parameter of the heating station can be automatically obtained through optimization only by giving a calibrated room temperature so as to control the heat supply of a heating part. Not only has high efficiency and shortest time consumption, but also can accurately reach the expected room temperature value of a user.
The control objective of the method is directly related to the indoor temperature of one or more rooms of the area supplied by the heat station, and the control logic is based on machine learning technology, especially artificial intelligence technology, and deep learning. In particular, a certain mathematical statistical treatment is carried out on the room temperature of the rooms to obtain a comprehensive quantity as a considerable amount. The selection of the room temperature is carried out by adopting a certain mathematical statistical model.
Compared with the prior art, the control scheme has the following innovation points:
1) the input parameter of the present control method may be a distribution or a distribution probability. In the traditional input parameters, the backwater temperature and the outdoor temperature are both determined values, and in the algorithm, the quantities can be distribution, and the expression form can be a distribution function, a data table and the like. The characteristic makes the algorithm adapt to the characteristic that the input quantity of the actual system is chaotic or probabilistic. In practical applications, such quantities are common in practice, and for example, the return water temperature, the outdoor temperature, and the like are actually a variable quantity.
2) Unlike the conventional method, the output decision quantity of the algorithm may not be a single value but may be a distribution of optimal values.
For example, when the algorithm is used for deciding the temperature of the secondary network supply water, the given decision output is not a single value, but the distribution of different optimal secondary network supply water temperatures, such as the expected value and the standard deviation of the normal distribution of the supply water temperature. The output capability of the algorithm can solve the problem that global optimization becomes possible when the control link is linked with other upper-layer optimization decision links.
3) The control objective of the method is directly linked to the indoor temperature of one or more rooms of the area supplied by the thermal station. Particularly, a comprehensive quantity is obtained as a considerable quantity after certain mathematical statistics processing is carried out on the room temperature of the rooms; and selecting the room temperatures by adopting a certain mathematical statistical model.
4) The control logic of the method is based on machine learning techniques, in particular artificial intelligence techniques, and in particular on deep learning.
5) The method for determining controllable tc is through mathematical optimization and planning, not logic, curve, table or formula. In particular on artificial intelligence techniques, and in particular on RL methods.
Embodiment 2 is an intelligent decision machine, which is an edge computing device, has an operating system built therein, and directly guides the control of the water supply temperature of a heat exchange station and a gas boiler on site based on the intelligent heat supply decision method described in embodiment 1, with the comprehensive room temperature of a user as a control target, thereby realizing the intelligent upgrade of a heat supply system.
The intelligent decision machine acquires data such as primary and secondary side flow, pressure, temperature, sampled room temperature, climate parameters and the like through the communication system and the cloud. The artificial intelligence AI built in the decision machine has the capabilities of logic deduction, rule identification and automatic optimization, can complete big data deep learning within 2-3 weeks, establishes a matched control model, and simultaneously feeds back and selects a control scheme in real time according to data to continuously evolve and give an optimal control parameter value.
The cloud end one-key operation after the product is put into operation, and the extremely simple back is a strong algorithm support: the decision machine AI can directly give out control target parameters such as water supply temperature and the like after completing complex operation according to the room temperature target data set by the user. The AI model of the decision machine can solve the problem of hysteresis of the room temperature data in the traditional control model, and combines climate parameters to predict and predict a reasonable control target value in advance, intervene in advance and stabilize the fluctuation of the room temperature.
The decision machine AI model can process a large amount of real-time data, excavate the energy consumption potential of the system from the data, give out a control mode beyond the traditional experience, and can further finely regulate and control, so that the energy-saving operation can be still realized even in the deep cold period. The data flow of the intelligent decision machine is as shown in the above figure, and the decision machine AI can complete functions of data modeling, deduction, optimization and the like, continuously evolve, continuously optimize the control effect, and exceed the traditional experience of people.
The decision machine has the following characteristics:
1. 4-core CPUs (central processing units) are configured, 4X1.5G/s are adopted, a plurality of decision machines support cluster type cooperation and parallel work for a multi-system heating station, and intelligent decision requirements of different data scales can be flexibly met.
2. A mathematical computation engine and an industrial Internet communication protocol stack are built in the system, the system is suitable for numerical computation and communication, and various open source applications and Python development languages are supported.
3. The built-in artificial intelligence engine is suitable for various edge calculation occasions needing artificial intelligence calculation, can complete deep learning of the heat exchange system model in only 2 weeks, and is put into an AI operation mode.
4. The debugging is simple, and the communication debugging of the operation data required by the decision machine is only required to be completed on site.
5. The intelligent control system perfectly replaces automatic control, realizes intelligent operation of the heating system without human intervention, achieves the constant temperature control effect when the amplitude of the average indoor temperature is less than 0.5 ℃, and is shown in figure 2. Before the decision machine is put into operation, the average room temperature of the system is about 18.4 ℃ for a long time. The temperature is set to be 19.5 ℃, the average room temperature rises to 18.8 ℃ 24 hours after the decision maker is put into operation, the average room temperature reaches 19.2 ℃ after 48 hours, the room temperature is stabilized at 19.7 ℃ after 72 hours, and the deviation from the set target value of 19.5 ℃ is +0.2 ℃. The operation is stable for 72 hours, and the following performance of the actual average room temperature and the set target room temperature is good under the condition that the fluctuation of the outdoor air temperature exceeds 10 ℃, and the deviation is less than 0.5 ℃.
6. The fluctuation range of the indoor average temperature from time to time is obviously reduced, the average indoor temperature can be reduced by 0.5 ℃ under the condition of ensuring the requirement of the lowest indoor temperature, and a new energy-saving space is created.
7. After the decision machine is adopted, the artificial intelligence AI regulates and controls the water outlet temperature of the intelligent decision machine set according to the target indoor temperature, and automatically sends instructions to the machine set to be executed, and the regulation frequency and the regulation time are finer.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present application is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
Although the invention has been described in detail above with reference to a general description and specific examples, it will be apparent to one skilled in the art that modifications or improvements may be made thereto based on the invention. Accordingly, such modifications and improvements are intended to be within the scope of the invention as claimed.
Claims (5)
1. A heat supply intelligent decision-making method based on artificial intelligence is characterized by comprising the following steps:
acquiring equivalent room temperature tn of at least one target space and a data set of a corresponding relation between the equivalent room temperature tn and a controllable quantity tc of a heating station at a heating position;
adopting an artificial intelligence algorithm, taking the controllable quantity tc of the thermal power station as an input parameter, taking the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter, and obtaining a prediction model P through training; taking a prediction model P as an Environment (Environment) of an reinforcement learning method, taking a controllable quantity tc of a thermal station as an Action (Action) of the reinforcement learning method, and taking a function value of which the absolute value of the difference between an equivalent room temperature tn and a target room temperature ts is in inverse proportion as a return (Reward) to train the prediction model P;
the controllable quantity tc of the heating power station is one or more of a primary side water supply temperature, a return water temperature, a primary side water supply pressure, a primary side electric regulating valve position, a water supply and return average temperature, a primary side flow, a primary side distributed pump frequency, a secondary side water supply temperature, a secondary side water supply pressure, a secondary side flow, a secondary side pump frequency, a return water pressure or heat of a heat supply position;
when the water supply temperature is decided in each control period, optimizing tc by taking the trained prediction model P as an object and the target room temperature ts as a target, finding the optimal tc, and outputting the tc as a controllable quantity target;
and after the optimal controllable quantity tc is determined, substituting the optimal controllable quantity tc into a functional relation formula tm = fm (tc), and deriving a physical controllable quantity tm from tc, wherein the physical controllable quantity tm is one or more of the derivative quantities of water supply temperature tg, water return temperature th, water supply and return weighted average temperature tp, flow Q and heat Q.
2. A heating intelligence decision making method based on artificial intelligence as claimed in claim 1 wherein a stochastic model is used to select the indoor temperature of at least one space.
3. A heating intelligence decision method based on artificial intelligence as claimed in claim 1 wherein the input parameter is a deterministic single value, a distribution function or a data table.
4. A heating intelligence decision method based on artificial intelligence as claimed in claim 1 wherein the output parameter is a deterministic single value, an optimal value distribution function or a data table.
5. An intelligent decision machine for an artificial intelligence based heating intelligent decision method according to any one of claims 1 to 4, comprising: the system comprises a CPU and one or more information collectors, wherein the information collectors are used for acquiring equivalent room temperature tn of a target space and controllable amount tc of a heating station at a heat supply position corresponding to the equivalent room temperature tn;
the CPU is used for obtaining a prediction model P by training by adopting an artificial intelligence algorithm and taking the controllable quantity tc of the thermal power station as an input parameter and the equivalent room temperature tn corresponding to the controllable quantity tc of the thermal power station as an output parameter; taking a prediction model P as an Environment (Environment) of an reinforcement learning method, taking a controllable quantity tc of a thermal station as an Action (Action) of the reinforcement learning method, and taking a function value of which the absolute value of the difference between an equivalent room temperature tn and a target room temperature ts is in inverse proportion as a return (Reward) to train the prediction model P;
the controllable quantity tc of the heating power station is one or more of a primary side water supply temperature, a return water temperature, a primary side water supply pressure, a primary side electric regulating valve position, a water supply and return average temperature, a primary side flow, a primary side distributed pump frequency, a secondary side water supply temperature, a secondary side water supply pressure, a secondary side flow, a secondary side pump frequency, a return water pressure or heat of a heat supply position;
and when the water supply temperature is decided in each control period, optimizing tc by taking the trained prediction model P as an object and the target room temperature ts as a target, finding the optimal tc, and outputting the tc as a controllable quantity target.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110426947.9A CN113110056B (en) | 2021-04-20 | 2021-04-20 | Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110426947.9A CN113110056B (en) | 2021-04-20 | 2021-04-20 | Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113110056A CN113110056A (en) | 2021-07-13 |
| CN113110056B true CN113110056B (en) | 2022-03-04 |
Family
ID=76719171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110426947.9A Active CN113110056B (en) | 2021-04-20 | 2021-04-20 | Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113110056B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115218268A (en) * | 2022-06-17 | 2022-10-21 | 史登峰 | Heating load output control method, heating load output control device, heating load output control apparatus, storage medium, and program product |
| CN115454156A (en) * | 2022-09-28 | 2022-12-09 | 北京市热力集团有限责任公司 | Intelligent boiler system based on double-linkage control and boiler control method |
| CN116699998B (en) * | 2023-06-21 | 2024-02-13 | 广州和达水务科技股份有限公司 | Unmanned on duty pressurization station control system based on AI algorithm |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523210A (en) * | 2020-04-10 | 2020-08-11 | 杭州英集动力科技有限公司 | Prediction analysis method and system for temperature rise and drop process of urban central heating system |
| CN111580382A (en) * | 2020-05-18 | 2020-08-25 | 瑞纳智能设备股份有限公司 | Unit-level heat supply adjusting method and system based on artificial intelligence |
| CN112594758A (en) * | 2020-11-21 | 2021-04-02 | 唐山曹妃甸热力有限公司 | Heat supply prediction method, device, equipment and storage medium |
-
2021
- 2021-04-20 CN CN202110426947.9A patent/CN113110056B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111523210A (en) * | 2020-04-10 | 2020-08-11 | 杭州英集动力科技有限公司 | Prediction analysis method and system for temperature rise and drop process of urban central heating system |
| CN111580382A (en) * | 2020-05-18 | 2020-08-25 | 瑞纳智能设备股份有限公司 | Unit-level heat supply adjusting method and system based on artificial intelligence |
| CN112594758A (en) * | 2020-11-21 | 2021-04-02 | 唐山曹妃甸热力有限公司 | Heat supply prediction method, device, equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113110056A (en) | 2021-07-13 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113110056B (en) | Heat supply intelligent decision-making method and intelligent decision-making machine based on artificial intelligence | |
| CN113657031B (en) | Heat supply dispatching automation realization method, system and platform based on digital twin | |
| CN113110057A (en) | Heating power station energy-saving control method based on artificial intelligence and intelligent decision system | |
| CN109270842B (en) | Bayesian network-based regional heat supply model prediction control system and method | |
| Korkas et al. | Grid-connected microgrids: Demand management via distributed control and human-in-the-loop optimization | |
| CN115220351B (en) | Intelligent energy-saving optimization control method for building air conditioning system based on cloud side end | |
| CN111561732B (en) | Heat exchange station heat supply adjusting method and system based on artificial intelligence | |
| CN112232980A (en) | Regulation and control method for heat pump unit of regional energy heat supply system | |
| CN111580382B (en) | Unit-level heat supply adjusting method and system based on artificial intelligence | |
| CN103591637A (en) | Centralized heating secondary network operation adjustment method | |
| CN112413762B (en) | Parameter optimization method and system for cooling water system of refrigerating room | |
| CN110260470A (en) | Central air-conditioning parallel connection cold load optimal distribution method based on colony intelligence framework | |
| CN113513783A (en) | Heat exchange station online monitoring system and heat exchange station control method | |
| Zhao et al. | Optimal control for hydraulic balance of secondary network in district heating system under distributed architecture | |
| CN117114290A (en) | Source network storage comprehensive energy interactive optimization system and method | |
| CN114777193B (en) | Method and system for switching household regulation and control modes of secondary network of heating system | |
| CN116436033A (en) | Temperature control load frequency response control method based on user satisfaction and reinforcement learning | |
| CN116592417A (en) | Centralized heating system optimal control method and system based on load prediction | |
| CN117114292A (en) | Multi-energy flow regulation and control method based on comprehensive energy system | |
| CN113609778B (en) | Multi-objective optimization method and system for comprehensive energy system | |
| Groumpos et al. | New advanced technology methods for energy efficiency of buildings | |
| CN116963461A (en) | Energy saving method and device for machine room air conditioner | |
| CN116147128A (en) | Energy-saving control method and system for heating ventilation air conditioner | |
| CN118153353B (en) | Dynamic simulation optimizing method and related device for air conditioner | |
| CN115051374A (en) | Control method and device for electric heating equipment to participate in electric power peak regulation and storage medium |
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 |