CN112923533B - Multi-agent-based hierarchical distributed optimization control method for central air-conditioning system - Google Patents
Multi-agent-based hierarchical distributed optimization control method for central air-conditioning system Download PDFInfo
- Publication number
- CN112923533B CN112923533B CN202110259769.5A CN202110259769A CN112923533B CN 112923533 B CN112923533 B CN 112923533B CN 202110259769 A CN202110259769 A CN 202110259769A CN 112923533 B CN112923533 B CN 112923533B
- Authority
- CN
- China
- Prior art keywords
- air
- intelligent
- conditioning box
- control
- conditioning
- 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
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/62—Control or safety arrangements characterised by the type of control or by internal processing, e.g. using fuzzy logic, adaptive control or estimation of values
- F24F11/63—Electronic processing
- F24F11/64—Electronic processing using pre-stored data
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F11/00—Control or safety arrangements
- F24F11/70—Control systems characterised by their outputs; Constructional details thereof
- F24F11/72—Control systems characterised by their outputs; Constructional details thereof for controlling the supply of treated air, e.g. its pressure
- F24F11/74—Control systems characterised by their outputs; Constructional details thereof for controlling the supply of treated air, e.g. its pressure for controlling air flow rate or air velocity
Abstract
The invention provides a hierarchical distributed optimal control method of a central air-conditioning system based on multiple intelligent agents, which solves the problem of distributed optimal control of a complex central air-conditioning system. The method decomposes a total optimization problem of an optimization control task of a central air conditioning system into a plurality of control levels by using a level alternating direction multiplier method, wherein each control level comprises solution of a series of sub-problems and update of dual variables, and the solution and the update are respectively realized by parallel local optimization calculation inside an intelligent agent and information interaction between the intelligent agents. The method takes the comfort degree of a terminal area and the total energy consumption of the central air-conditioning system as optimization targets, and the optimal operation strategy of the central air-conditioning system is calculated through a hierarchical distributed optimization control framework. The method adopts the idea of asynchronous calculation, has higher calculation speed and can meet the real-time control requirement. Meanwhile, the hierarchical distributed framework provided by the method has strong expandability and flexibility.
Description
Technical Field
The invention relates to the technical field of building energy conservation and intelligent building control, in particular to a hierarchical distributed optimization control method of a central air-conditioning system based on a multi-agent technology.
Background
The central air-conditioning system is widely applied to the field of buildings, and the energy consumption of the central air-conditioning system accounts for a large proportion of the total energy consumption of the buildings. Therefore, the optimization control strategy of the central air-conditioning system is researched, so that the energy consumption can be reduced as far as possible on the premise of meeting the requirement of the terminal dynamic load, and the optimization control strategy has very important significance for realizing building energy conservation.
The optimal control method of the central air-conditioning system may be divided into centralized control and distributed control. The distributed control method has the characteristics of high flexibility, high expansibility, high control efficiency and the like, and has become a hotspot of research in the field of optimization control of central air-conditioning systems in recent years. The existing distributed control method usually only focuses on the control strategy optimization of a certain subsystem in the central air-conditioning system. For example, the patent "central air-conditioning host operation optimization control system and method based on distributed computation" realizes the distributed optimization control of the number of water chilling unit groups and the cooling capacity; the patent 'a group intelligent optimization method for dynamic hydraulic balance of a chilled water pipe network of a central air conditioner' realizes distributed optimization control of chilled water pump groups and the hydraulic power of the chilled water pipe network in a chilled water system; the patent 'centerless self-organizing variable air volume terminal optimization control method and system' realizes distributed optimization control of the variable air volume air supply system. In fact, a typical central air conditioning system comprises a water system, an air supply system and corresponding air supply terminals, the optimization of the system involves simultaneously realizing the energy saving of the system and the comfort of each terminal room, and due to the strong coupling of the thermal processes among subsystems, the situation of 'this trade-off' exists in the optimization control effect, and the global optimization is difficult to realize. The existing distributed control strategy only focuses on reducing the energy consumption of some equipment in the system or realizing the comfort of end rooms, is difficult to apply to the distributed optimization control of the whole central air-conditioning system, and still lacks an effective solution at present.
Disclosure of Invention
The invention aims to overcome the defects of the distributed control method of the existing central air-conditioning system, and provides an optimized control method of the central air-conditioning system with distributed intelligence based on an Alternating Direction Method of Multipliers (ADMM). The method divides the overall optimization goal of the central air-conditioning system into a plurality of levels through a level ADMM algorithm. Taking a two-layer control structure as an example, the upper layer optimization realizes the energy saving of the system by optimizing the set value parameters of the equipment, the lower layer optimization realizes the maintenance of the comfort level of the air supply area by optimizing the set value of the parameters at the tail end of the air supply, and the upper layer and the lower layer realize the optimization of the overall target by continuous iteration. The method is characterized in that three types of intelligent agents are arranged, including a coordination intelligent agent, an equipment intelligent agent (including a water chilling unit intelligent agent and an air conditioning box intelligent agent) and a room intelligent agent, and the hierarchical distributed control task is realized through local optimization functions of the three types of intelligent agents and information interaction among the three types of intelligent agents. The method takes the comfort level of a terminal area to be maintained and the total energy consumption of the central air-conditioning system to be reduced as optimization targets, takes the energy conservation among equipment and the air volume conservation of an air-conditioning box as constraints, calculates the optimal operation strategy of the central air-conditioning system through a hierarchical distributed optimization control framework, effectively solves the distributed optimization control of the central air-conditioning system level, and has strong flexibility and expandability.
The invention adopts the following technical scheme:
a hierarchical distributed optimization control method of a central air-conditioning system based on multiple intelligent agents is characterized by comprising the following steps:
s1: respectively setting a corresponding intelligent agent for each water chilling unit, each air conditioning box and each air conditioning room connected with the air conditioning box in the central air conditioning system;
s2: setting initial parameters, wherein the specific steps are as follows from S2-1 to S2-3:
s2-1: setting an initial chilled water supply water temperature set point
Initial air volume of air conditioning cabinet
And initial dual variables
S2-2: calculating the cold load value Q of each roomij；
S2-3: setting an optimal control interval T;
s3: issuing a hierarchical distributed optimization control task, and setting a total optimization target as follows:
wherein N ischillerNumber of water chilling units, NAHUNumber of air-conditioning boxes, NAHUiThe number of air-conditioning rooms connected with the ith air-conditioning box is omega, which is a weight factor; pm(. h) is the power of the mth water chilling unit, and the chilled water is used for supplying water with a set value TchwsDetermining; pi(. h) is the power of the ith air-conditioning box, and the air quantity q is delivered by the ith air-conditioning boxAHUiDetermining; cij(. h) comfort level of j th room connected to i-th air-conditioning box, j th variable air volume end air volume q connected to the air-conditioning boxijDetermining; s.t. formula notes the constraints of the optimization problem; the first type of constraint is the conservation of energy between the heat exchange capacity of the air conditioning cabinet and the cooling load of the air conditioning room and the heat dissipation of the fan, wherein HXi(. h) is the heat exchange quantity of the ith air-conditioning box, from TchwsAnd q isAHUiDetermining; qijThe cold load value of the jth room corresponding to the ith air-conditioning box is obtained; the second type of constraint is that the quality between the total air supply quantity of the air conditioning box and the tail air supply quantities of all the variable air quantity connected with the air conditioning box is constant, qijThe air supply value of the tail end of the jth room corresponding to the ith air-conditioning box is obtained;
s4: decomposing a total optimization control task of the central air-conditioning system into a plurality of control levels by using a level alternating direction multiplier method, wherein each control level comprises the solution of a series of sub-problems and the update of dual variables; in each iterative calculation, the upper control level downloads the updated values of the local optimization variables and the dual variables step by step, and the lower control level executes the optimization task and then sends the optimization results back to the upper control level step by step; realizing distributed optimization control of the central air-conditioning system through asynchronous iterative computation among the levels;
s5: and when the control time interval reaches T, repeating the steps S2-S4 to realize real-time optimal control.
Preferably, in step S4, for the central air-conditioning system, a double-layer alternating direction multiplier method is used to implement distributed optimal control of the central air-conditioning system; in the upper-layer control, the intelligent water chilling unit and the intelligent air conditioning box solve corresponding sub-problems in parallel, and information interaction between the intelligent agents of the equipment layer is realized through the coordination of the intelligent agents; in the lower-layer control, the room intelligent agents solve the corresponding sub-problems in parallel, and the information interaction between the room-level intelligent agents is realized through the air-conditioning box intelligent agents connected with the room intelligent agents; the distributed optimal control of the central air-conditioning system is finally realized through asynchronous iterative calculation of upper-layer control and lower-layer control; the method comprises the following specific steps in sequence:
s4-1: coordinating the agent to send the initial chilled water supply temperature set value to all the chiller agent agents and the air conditioning box agents
Initial air volume of air conditioning cabinet
And initial dual variables
S4-2: setting the upper iteration number ku0 and the maximum iteration number K of the upper layeru,max；
S4-3: each water chilling unit intelligent body receives the variable value sent by the coordination intelligent body and carries out chilled water supply temperature set value by using the following formula
Updates the computation asynchronously and sends back to the coordinating agent:
where ρ isuA penalty factor being a first type of constraint; superscript k of parametersuAll represent the kuThe next upper layer iteration of corresponding ginsengCounting;
s4-4: each air-conditioning box intelligent agent receives the variable value sent by the coordination intelligent agent, and sets the lower layer iteration times kl,i0 and the lower maximum number of iterations Kl,max(ii) a Setting initial air-conditioning box air supply amount for each air-conditioning box intelligent body
Initial variable air volume end air volume
And initial dual variables
And transmitting the initial values to each room agent connected with the room agent;
s4-5: each room intelligent body receives the variable value sent by the corresponding air conditioning box intelligent body, and the variable air volume tail end air supply volume is carried out according to the following formula
And (3) updating and calculating in an asynchronous manner, and sending back the corresponding air conditioning box intelligent agent:
where ρ islA penalty factor for a second type of constraint; superscript k of parametersl,iAll represent the kl,iThe parameters corresponding to the next upper layer iteration;
s4-6: the air supply quantity of the air conditioning box is carried out by each intelligent air conditioning box according to the following formula
The update calculation of (2):
in the formula: p (-) is the power of the air conditioning box;
s4-7: each air-conditioning box intelligent agent is used for receiving the variable air volume tail end air supply volume
And air volume of air-conditioning case
Judging whether the convergence conditions of the lower algorithms are met; if so, the intelligent air-conditioning box body sends the optimal air-conditioning box air supply quantity value qAHUiSending the information to a coordination agent, and jumping to S4-8; if not, the dual variables are performed according to the following formula
And sends the values back to all corresponding room agents, and then sets the iteration number kl,iIncreasing the value by 1, and jumping to S4-5;
s4-8: the coordination intelligent agent supplies water temperature set value according to the received chilled water
And air volume of air-conditioning case
Judging whether the convergence condition of the upper algorithm is met; if yes, ending the current hierarchical distributed optimization control task; if not, the dual variables are performed according to the following formula
And sending the values back to all the intelligent water chilling unit bodies and the intelligent air conditioning box bodies, and then setting the iteration number kuIncreasing the value by 1, and jumping to S4-3;
preferably, the initial values of each optimized variable and the dual variable in the upper layer and the lower layer are set as the iteration final values of the algorithm in the last time interval.
Preferably, the optimal control interval T is determined based on the sensor data sampling interval.
Preferably, the condition for the coordination agent to judge convergence is that the following three formulas are simultaneously satisfied or the iteration number kuThe maximum iteration number K of the upper layer is reachedu,max：
Wherein σu1，σu2，σu3Is a threshold value.
Preferably, the condition for each air-conditioning box agent to judge convergence is that the following three formulas are simultaneously satisfied or the iteration number kl,iUp to the maximum number of iterations Kl,max：
Wherein σl1，σl2Is the corresponding threshold.
Compared with the prior art, the invention has at least the following beneficial effects:
(1) the distributed optimization control method of the central air-conditioning system based on the hierarchical alternating direction multiplier method divides the overall optimization target of the central air-conditioning system into an upper layer optimization layer and a lower layer optimization layer, wherein the upper layer optimization realizes the energy conservation of the central air-conditioning system equipment, the lower layer optimization realizes the maintenance of the comfort level of an air supply area, the hierarchical distributed optimization control of the central air-conditioning system is realized through repeated iterative calculation between the upper layer and the lower layer, and the global optimization capability is realized.
(2) The hierarchical alternative direction multiplier method provided by the invention adopts the idea of asynchronous calculation, and the optimization of the upper layer and the lower layer can have high-speed parallel calculation capability as the traditional alternative direction multiplier method. In each iterative calculation process, local optimization tasks can be simultaneously and parallelly executed between the intelligent water chilling unit and each intelligent air conditioning box and between the intelligent air conditioning box and the corresponding intelligent variable air volume tail end, the calculation speed of the hierarchical alternating direction multiplier method is greatly increased, and the real-time control capability is good when the control is carried out on a complex central air conditioning system.
Drawings
FIG. 1 is a schematic diagram of a central air conditioning system according to an embodiment of the present invention;
FIG. 2 is a communication process between a hierarchy and agents in an embodiment of the present invention;
FIG. 3 is a flowchart illustrating a method for calculating a hierarchical alternative direction multiplier according to an embodiment of the present invention;
Detailed Description
The invention will be further elucidated and described with reference to the drawings and examples.
In a preferred embodiment of the present invention, a hierarchical distributed optimization control method for a multi-agent-based central air conditioning system is provided, which comprises the following steps:
s1: and respectively setting a corresponding intelligent agent for each water chilling unit, each air conditioning box and each air conditioning room connected with the air conditioning box in the central air conditioning system. As shown in fig. 1, in this embodiment, for 2 chiller units, 2 air-conditioning boxes and 6 air-conditioning rooms connected to the chiller units and the air-conditioning boxes in the central air-conditioning system, a corresponding agent is respectively provided, that is, 2 chiller unit agents, 2 air-conditioning box agents and 6 room agents are respectively provided.
The communication process between the hierarchical architecture and the agents in this embodiment is shown in fig. 2, the calculation flow of the hierarchical alternative direction multiplier method in this embodiment is shown in fig. 3, and the specific implementation process of the method is described below.
S2: setting initial parameters, wherein the specific steps are as follows from S2-1 to S2-3:
s2-1: setting an initial chilled water supply water temperature set point
Initial air volume of air conditioning cabinet
And initial dual variables
S2-2: calculating the cold load value Q of each roomij；
S2-3: an optimal control interval T is set. The optimal control interval T may be determined based on the sensor data sampling interval.
S3: issuing a hierarchical distributed optimization control task, and setting a total optimization target as follows:
wherein N ischillerNumber of water chilling units, NAHUNumber of air-conditioning boxes, NAHUiIs an ith stationThe number of air-conditioning rooms connected with the air-conditioning box is omega which is a weight factor; pm(. h) is the power of the mth water chilling unit, and the chilled water is used for supplying water with a set value TchwsDetermining; pi(. h) is the power of the ith air-conditioning box, and the air quantity q is delivered by the ith air-conditioning boxAHUiDetermining; cij(. h) comfort level of j th room connected to i-th air-conditioning box, j th variable air volume end air volume q connected to the air-conditioning boxijDetermining; s.t. formula notes the constraints of the optimization problem; the first type of constraint is the conservation of energy between the heat exchange capacity of the air conditioning cabinet and the cooling load of the air conditioning room and the heat dissipation of the fan, wherein HXi(. h) is the heat exchange quantity of the ith air-conditioning box, from TchwsAnd q isAHUiDetermining; qijThe cold load value of the jth room corresponding to the ith air-conditioning box is obtained; the second type of constraint is that the quality between the total air supply quantity of the air conditioning box and the tail air supply quantities of all the variable air quantity connected with the air conditioning box is constant, qijAnd the air supply value is the tail end air supply value of the jth room corresponding to the ith air conditioning box.
S4: decomposing a total optimization control task of the central air-conditioning system into a plurality of control levels by using a level alternating direction multiplier method, wherein each control level comprises the solution of a series of sub-problems and the update of dual variables; in each iterative calculation, the upper control level downloads the updated values of the local optimization variables and the dual variables step by step, and the lower control level executes the optimization task and then sends the optimization results back to the upper control level step by step; and realizing distributed optimization control of the central air-conditioning system through asynchronous iterative computation among the levels.
In the embodiment, the central air-conditioning system is divided into three levels, namely a system level, an equipment level and a room level, and the overall optimization problem of the optimization control task of the central air-conditioning system is decomposed into two control levels by using a level alternating direction multiplier method. For the central air-conditioning system, the distributed optimal control of the central air-conditioning system is realized by adopting a double-layer alternative direction multiplier method; in the upper-layer control, the intelligent water chilling unit and the intelligent air conditioning box solve corresponding sub-problems in parallel, and information interaction between the intelligent agents of the equipment layer is realized through the coordination of the intelligent agents; in the lower-layer control, the room intelligent agents solve the corresponding sub-problems in parallel, and the information interaction between the room-level intelligent agents is realized through the air-conditioning box intelligent agents connected with the room intelligent agents; and finally realizing the distributed optimal control of the central air-conditioning system through asynchronous iterative calculation of the upper-layer control and the lower-layer control. Therefore, the specific steps of S4 are:
s4-1: coordinating the agent to send the initial chilled water supply temperature set value to all the chiller agent agents and the air conditioning box agents
Initial air volume of air conditioning cabinet
And initial dual variables
S4-2: setting the upper iteration number ku0 and the maximum iteration number K of the upper layeru,max；
S4-3: each water chilling unit intelligent body receives the variable value sent by the coordination intelligent body and carries out chilled water supply temperature set value by using the following formula
Updates the computation asynchronously and sends back to the coordinating agent:
where ρ isuA penalty factor being a first type of constraint; superscript k of parametersuAll represent the kuThe parameters corresponding to the next upper layer iteration;
s4-4: each air-conditioning box intelligent agent receives the variable value sent by the coordination intelligent agent, and sets the lower layer iteration times kl,i0 and the lower maximum number of iterations Kl,max(ii) a Setting initial air-conditioning box air supply amount for each air-conditioning box intelligent body
Initial variable air volume end air volume
And initial dual variables
And transmitting the initial values to each room agent connected with the room agent;
s4-5: each room intelligent body receives the variable value sent by the corresponding air conditioning box intelligent body, and the variable air volume tail end air supply volume is carried out according to the following formula
And (3) updating and calculating in an asynchronous manner, and sending back the corresponding air conditioning box intelligent agent:
where ρ islA penalty factor for a second type of constraint; superscript k of parametersl,iAll represent the kl,iThe parameters corresponding to the next upper layer iteration;
s4-6: the air supply quantity of the air conditioning box is carried out by each intelligent air conditioning box according to the following formula
The update calculation of (2):
in the formula: p (-) is the power of the air conditioning box;
s4-7: each air-conditioning box intelligent agent is used for receiving the variable air volume tail end air supply volume
And air volume of air-conditioning case
Judging whether the convergence conditions of the lower algorithms are met; if so, the intelligent air-conditioning box body sends the optimal air-conditioning box air supply quantity value qAHUiSending the information to a coordination agent, and jumping to S4-8; if not, the dual variables are performed according to the following formula
And sends the values back to all corresponding room agents, and then sets the iteration number kl,iIncreasing the value by 1, and jumping to S4-5;
s4-8: the coordination intelligent agent supplies water temperature set value according to the received chilled water
And air volume of air-conditioning case
Judging whether the convergence condition of the upper algorithm is met; if yes, ending the current hierarchical distributed optimization control task; if not, the dual variables are performed according to the following formula
And sending the values back to all the intelligent water chilling unit bodies and the intelligent air conditioning box bodies, and then setting the iteration number kuIncreasing the value by 1, and jumping to S4-3;
in this embodiment, the initial values of the optimization variables and the dual variables in the upper layer and the lower layer control may be set as the iteration final values of the previous time interval algorithm.
In this embodiment, the condition for coordinating the judgment convergence of the agent is that the following three formulas are simultaneously satisfied or the number k of iterationsuThe maximum iteration number K of the upper layer is reachedu,max。
Wherein σu1，σu2，σu3Is the corresponding threshold.
In this embodiment, the condition for each air-conditioning box agent to judge convergence is that the following three formulas are simultaneously satisfied or the iteration number kl,iUp to the maximum number of iterations Kl,max。
Wherein σl1，σl2Is the corresponding threshold.
S5: and when the control time interval reaches T, repeating the steps S2-S4 to realize real-time optimal control.
In the embodiment, the optimization problem of the complex central air-conditioning system is subjected to hierarchical decomposition and hierarchical calculation by adopting a hierarchical distributed calculation method, the hierarchical decomposition and the hierarchical calculation are divided into an upper layer optimization layer and a lower layer optimization layer, the upper layer optimization layer realizes energy conservation of the central air-conditioning system equipment, the lower layer optimization layer maximizes comfort level indexes of each air supply area, the intelligent bodies corresponding to the upper layer and the lower layer continuously perform iterative information exchange, and finally the optimal chilled water supply temperature set value, the optimal air supply quantity set value of each air-conditioning box and the optimal air supply quantity set value of each variable air quantity tail end of each water chilling unit at each time are obtained.
In the above embodiment, the hierarchical alternative direction multiplier method used adopts the idea of asynchronous calculation, and the upper and lower layer optimization can have high-speed parallel computing capability as the conventional alternative direction multiplier method. In each iterative calculation process, local optimization tasks can be simultaneously and parallelly executed between the intelligent water chilling unit and each intelligent air conditioning box and between the intelligent air conditioning box and the corresponding intelligent variable air volume tail end, the calculation speed of the hierarchical alternating direction multiplier method is greatly increased, and the real-time control capability is good when the control is carried out on a complex central air conditioning system.
The above-described embodiments are merely preferred embodiments of the present invention, which should not be construed as limiting the invention. Various changes and modifications, particularly in the use of different control levels, may also be made by those of ordinary skill in the pertinent art without departing from the spirit and scope of the present invention. Therefore, the technical scheme obtained by adopting the mode of equivalent replacement or equivalent transformation is within the protection scope of the invention.
Claims (6)
1. A hierarchical distributed optimization control method of a central air-conditioning system based on multiple intelligent agents is characterized by comprising the following steps:
s1: respectively setting a corresponding intelligent agent for each water chilling unit, each air conditioning box and each air conditioning room connected with the air conditioning box in the central air conditioning system;
s2: setting initial parameters, wherein the specific steps are as follows from S2-1 to S2-3:
s2-1: setting an initial chilled water supply water temperature set point
Initial air volume of air conditioning cabinet
And initial dual variables
S2-2: calculating the cold load value Q of each roomij；
S2-3: setting an optimal control interval T;
s3: issuing a hierarchical distributed optimization control task, and setting a total optimization target as follows:
wherein N ischillerNumber of water chilling units, NAHUNumber of air-conditioning boxes, NAHUiThe number of air-conditioning rooms connected with the ith air-conditioning box is omega, which is a weight factor; pm(. h) is the power of the mth water chilling unit, and the chilled water is used for supplying water with a set value TchwsDetermining; pi(. h) is the power of the ith air-conditioning box, and the air quantity q is delivered by the ith air-conditioning boxAHUiDetermining; cij(. h) comfort level of j th room connected to i-th air-conditioning box, j th variable air volume end air volume q connected to the air-conditioning boxijDetermining; s.t. formula notes the constraints of the optimization problem; the first type of constraint is the conservation of energy between the heat exchange capacity of the air conditioning cabinet and the cooling load of the air conditioning room and the heat dissipation of the fan, wherein HXi(. h) is the heat exchange quantity of the ith air-conditioning box, from TchwsAnd q isAHUiDetermining; qijThe cold load value of the jth room corresponding to the ith air-conditioning box is obtained; the second type of constraint is that the total air supply volume of the air conditioning box is connected with each air conditioning boxConstant quality between variable air volume and end air volume, qijThe air supply value of the tail end of the jth room corresponding to the ith air-conditioning box is obtained;
s4: decomposing a total optimization control task of the central air-conditioning system into a plurality of control levels by using a level alternating direction multiplier method, wherein each control level comprises the solution of a series of sub-problems and the update of dual variables; in each iterative calculation, the upper control level downloads the updated values of the local optimization variables and the dual variables step by step, and the lower control level executes the optimization task and then sends the optimization results back to the upper control level step by step; realizing distributed optimization control of the central air-conditioning system through asynchronous iterative computation among the levels;
s5: and when the control time interval reaches T, repeating the steps S2-S4 to realize real-time optimal control.
2. The multi-agent based hierarchical distributed optimization control method of a central air conditioning system as claimed in claim 1, characterized in that: in the step S4, for the central air-conditioning system, a double-layer alternating direction multiplier method is used to realize distributed optimal control of the central air-conditioning system; in the upper-layer control, the intelligent water chilling unit and the intelligent air conditioning box solve corresponding sub-problems in parallel, and information interaction between the intelligent agents of the equipment layer is realized through the coordination of the intelligent agents; in the lower-layer control, the room intelligent agents solve the corresponding sub-problems in parallel, and the information interaction between the room-level intelligent agents is realized through the air-conditioning box intelligent agents connected with the room intelligent agents; the distributed optimal control of the central air-conditioning system is finally realized through asynchronous iterative calculation of upper-layer control and lower-layer control; the method comprises the following specific steps in sequence:
s4-1: coordinating the agent to send the initial chilled water supply temperature set value to all the chiller agent agents and the air conditioning box agents
Initial air volume of air conditioning cabinet
And initial dual variables
S4-2: setting the upper iteration number ku0 and the maximum iteration number K of the upper layeru,max；
S4-3: each water chilling unit intelligent body receives the variable value sent by the coordination intelligent body and carries out chilled water supply temperature set value by using the following formula
Updates the computation asynchronously and sends back to the coordinating agent:
where ρ isuA penalty factor being a first type of constraint; superscript k of parametersuAll represent the kuThe parameters corresponding to the next upper layer iteration;
s4-4: each air-conditioning box intelligent agent receives the variable value sent by the coordination intelligent agent, and sets the lower layer iteration times kl,i0 and the lower maximum number of iterations Kl,max(ii) a Setting initial air-conditioning box air supply amount for each air-conditioning box intelligent body
Initial variable air volume end air volume
And initial dual variables
And transmitting the initial values to each room agent connected with the room agent;
s4-5: each room intelligent body receives the variable value sent by the corresponding air conditioning box intelligent body and carries out the variable value according to the following formulaVariable air volume end air volume
And (3) updating and calculating in an asynchronous manner, and sending back the corresponding air conditioning box intelligent agent:
where ρ islA penalty factor for a second type of constraint; superscript k of parametersl,iAll represent the kl,iThe parameters corresponding to the next upper layer iteration;
s4-6: the air supply quantity of the air conditioning box is carried out by each intelligent air conditioning box according to the following formula
The update calculation of (2):
in the formula: p (-) is the power of the air conditioning box;
s4-7: each air-conditioning box intelligent agent is used for receiving the variable air volume tail end air supply volume
And air volume of air-conditioning case
Judging whether the convergence conditions of the lower algorithms are met; if so, the intelligent air-conditioning box body sends the optimal air-conditioning box air supply quantity value qAHUiSending the information to a coordination agent, and jumping to S4-8; if not, the dual variables are performed according to the following formula
And sends the above values back to all corresponding room agents, after which the settings are madeNumber of iterations kl,iIncreasing the value by 1, and jumping to S4-5;
s4-8: the coordination intelligent agent supplies water temperature set value according to the received chilled water
And air volume of air-conditioning case
Judging whether the convergence condition of the upper algorithm is met; if yes, ending the current hierarchical distributed optimization control task; if not, the dual variables are performed according to the following formula
And sending the values back to all the intelligent water chilling unit bodies and the intelligent air conditioning box bodies, and then setting the iteration number kuIncreasing the value by 1, and jumping to S4-3;
3. the multi-agent based hierarchical distributed optimization control method of a central air conditioning system as claimed in claim 1, characterized in that: initial values of optimized variables and dual variables in the upper layer control and the lower layer control are set as iteration final values of the last time interval algorithm.
4. The multi-agent based hierarchical distributed optimization control method of a central air conditioning system as claimed in claim 1, characterized in that: the optimal control interval T is determined from the sensor data sampling interval.
5. As claimed in claim 1The hierarchical distributed optimization control method of the central air-conditioning system based on the multi-agent is characterized by comprising the following steps: the condition for coordinating the judgment convergence of the intelligent agent is that the following three formulas are simultaneously satisfied or the iteration number kuThe maximum iteration number K of the upper layer is reachedu,max：
Wherein σu1，σu2，σu3Is a threshold value.
6. The multi-agent based hierarchical distributed optimization control method of a central air conditioning system as claimed in claim 1, characterized in that: the condition for judging convergence by each air conditioning box intelligent agent is that the following three formulas are simultaneously met or the iteration number kl,iUp to the maximum number of iterations Kl,max：
Wherein σl1，σl2Is the corresponding threshold.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110259769.5A CN112923533B (en) | 2021-03-10 | 2021-03-10 | Multi-agent-based hierarchical distributed optimization control method for central air-conditioning system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110259769.5A CN112923533B (en) | 2021-03-10 | 2021-03-10 | Multi-agent-based hierarchical distributed optimization control method for central air-conditioning system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112923533A CN112923533A (en) | 2021-06-08 |
| CN112923533B true CN112923533B (en) | 2021-11-30 |
Family
ID=76172381
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110259769.5A Active CN112923533B (en) | 2021-03-10 | 2021-03-10 | Multi-agent-based hierarchical distributed optimization control method for central air-conditioning system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112923533B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116681269B (en) * | 2023-08-03 | 2023-10-13 | 南京邮电大学 | Intelligent collaborative operation optimization method for power grid interactive type efficient residential building |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102084316A (en) * | 2008-05-05 | 2011-06-01 | 西门子工业公司 | Arrangement for managing data center operations to increase cooling efficiency |
| JP2012137246A (en) * | 2010-12-27 | 2012-07-19 | Mitsubishi Electric Corp | Air conditioner system |
| JP2014240729A (en) * | 2013-06-12 | 2014-12-25 | 株式会社東芝 | Air conditioning energy management system, air conditioning energy management method, and program |
| CN109084403A (en) * | 2018-06-29 | 2018-12-25 | 广州能迪能源科技股份有限公司 | Water cooler static cost control strategy preparation method based on air conditioner load timing distribution |
-
2021
- 2021-03-10 CN CN202110259769.5A patent/CN112923533B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102084316A (en) * | 2008-05-05 | 2011-06-01 | 西门子工业公司 | Arrangement for managing data center operations to increase cooling efficiency |
| JP2012137246A (en) * | 2010-12-27 | 2012-07-19 | Mitsubishi Electric Corp | Air conditioner system |
| JP2014240729A (en) * | 2013-06-12 | 2014-12-25 | 株式会社東芝 | Air conditioning energy management system, air conditioning energy management method, and program |
| CN109084403A (en) * | 2018-06-29 | 2018-12-25 | 广州能迪能源科技股份有限公司 | Water cooler static cost control strategy preparation method based on air conditioner load timing distribution |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112923533A (en) | 2021-06-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110288164B (en) | Predictive control method for building air-conditioning refrigeration station system | |
| CN111256294B (en) | Model prediction-based optimization control method for combined operation of water chilling unit | |
| CN108413567B (en) | Central air conditioner cost-saving optimization method and system based on Internet of things | |
| CN111561732B (en) | Heat exchange station heat supply adjusting method and system based on artificial intelligence | |
| CN110392515A (en) | A kind of Cooling and Heat Source equipment room energy-conserving control method and system based on historical data | |
| CN108446809A (en) | A kind of regional complex energy device and network dual-layer optimization configuration method | |
| CN101655272A (en) | Energy-saving control management system of network central air conditioner and method thereof | |
| CN112923533B (en) | Multi-agent-based hierarchical distributed optimization control method for central air-conditioning system | |
| CN110260470A (en) | Central air-conditioning parallel connection cold load optimal distribution method based on colony intelligence framework | |
| CN106600080A (en) | Data network and power network coupling model participation system economic dispatch method | |
| CN112344522B (en) | Load distribution type optimal configuration method for central air-conditioning cooler system | |
| CN112292013A (en) | Micro-module data center heat source prejudgment and cold source regulation and control method based on time sequence | |
| Zhuang et al. | A decentralized method for energy conservation of an HVAC system | |
| CN111415036B (en) | Load optimization distribution method for parallel connection cold machines of central air-conditioning system | |
| CN107101324B (en) | A kind of consistency control method of convertible frequency air-conditioner load group | |
| Liu et al. | A decentralized optimization method for energy saving of HVAC systems | |
| CN113064480A (en) | Multi-data center collaborative energy saving method based on multi-agent reinforcement learning | |
| CN113028613B (en) | Distributed optimization control method of variable air volume system based on rapid ADMM | |
| CN111219856B (en) | Air treatment equipment intelligent optimization group control device and method based on 5G communication | |
| CN113685971A (en) | Constant-temperature heat supply automatic control circulation dynamic balance control system and method and data processing terminal | |
| Endo et al. | Development of predictive control system using just-in-time modeling and enthalpy-aware control in air conditioners for large-scale data center | |
| CN112432269A (en) | Method and system for optimizing set value of pressure difference of refrigerating water pump of refrigerating room | |
| CN114115398B (en) | Demand response control method and device for building water chilling unit | |
| Lin et al. | Optimizing for Large Time Delay Systems by BP Neural Network and Evolutionary Algorithm Improving. | |
| CN115470998B (en) | Port cold box load group power consumption consistency layering optimization scheduling method and 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 |