CN113418288A - Simulation model-based neural network multi-terminal air valve control system and method - Google Patents
Simulation model-based neural network multi-terminal air valve control system and method Download PDFInfo
- Publication number
- CN113418288A CN113418288A CN202110746780.4A CN202110746780A CN113418288A CN 113418288 A CN113418288 A CN 113418288A CN 202110746780 A CN202110746780 A CN 202110746780A CN 113418288 A CN113418288 A CN 113418288A
- Authority
- CN
- China
- Prior art keywords
- air
- variable
- air volume
- air valve
- control
- 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
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
- F24F11/70—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
- F24F11/77—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 by controlling the speed of ventilators
-
- 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
-
- 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
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F24—HEATING; RANGES; VENTILATING
- F24F—AIR-CONDITIONING; AIR-HUMIDIFICATION; VENTILATION; USE OF AIR CURRENTS FOR SCREENING
- F24F2110/00—Control inputs relating to air properties
- F24F2110/10—Temperature
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02B—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO BUILDINGS, e.g. HOUSING, HOUSE APPLIANCES OR RELATED END-USER APPLICATIONS
- Y02B30/00—Energy efficient heating, ventilation or air conditioning [HVAC]
- Y02B30/70—Efficient control or regulation technologies, e.g. for control of refrigerant flow, motor or heating
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Fluid Mechanics (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Signal Processing (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention provides a neural network multi-terminal air valve control system and method based on a simulation model, which relate to the technical field of air conditioning and ventilation energy-saving control and comprise a control host, wherein the control host is connected with a variable frequency fan through a fan frequency controller, and the variable frequency fan is connected with a plurality of variable air volume air valves; a plurality of variable air volume air valves are respectively connected through a plurality of electric air valve actuators; the air conditioning rooms are all provided with a temperature sensor, a temperature control panel and a temporary air volume sensor and are connected to the control host; the control host stores an air pipe system simulation model and an optimization control program, the air pipe system simulation model provides training data for an air pipe model learning module in the optimization control program, and the air pipe model learning module trains to obtain a neural network data model; and the neural network data model calculates the opening degree of each variable air volume air valve according to the data of each element and controls the opening degrees of the variable air volume air valves. The invention improves the control precision of the opening degree of the multi-tail-end variable air volume air valve.
Description
Technical Field
The invention relates to the technical field of air conditioning and ventilation energy-saving control, in particular to a neural network multi-terminal air valve control system and method based on a simulation model.
Background
The air-conditioning system with variable air volume is widely applied to the field of building air-conditioning due to the flexibility of control, and the opening degree of an air valve in the air-conditioning system with variable air volume is a direct factor influencing the air volume at each tail end. The air volume control of the variable air volume air conditioning system usually adopts two control methods: (1) the tail ends are controlled by fixed gears, the opening degree of the air valve is set to be a fixed full-open/full-close two-gear, or fixed gear numbers (such as close, 30% opening degree, 50% opening degree, 70% opening degree and full-open five-gear) are selected according to the opening degree, the opening degree of the air valve is determined according to the gears, the opening degree change number is reduced, and the fan control is simplified. (2) And PID controls the fan frequency to enable the total air inlet volume of the air conditioning system to meet the requirement, the air volume of each tail end is adjusted through the opening of an air valve, and the opening of the air valve is subjected to PID regulation and control according to the set real-time temperature of a room. (3) And controlling the opening of the air valve through a big data model, and determining the opening of the air valve under different working conditions based on field historical data.
At present, the method for controlling the variable air volume by adopting the fixed gear is applied more in practical situations, and has the advantages of simple operation, easy implementation and the defects of rough control and lower precision. The method for controlling the fan frequency by the PID focuses on local optimization, the air valve at the tail end of each variable air volume air conditioning system only performs PID regulation and control on a single tail end, but the local optimization is not always global optimal under the condition of multiple tail ends, and the coupling effect among the air valves can cause the condition that even if the opening degree of partial tail ends is fully opened, the air supply requirement cannot be met. Controlling the dampers through a big data model helps to solve the coupling problem between dampers, but puts a great demand on the amount of training data. In most cases, sufficient field data are lacked, so that a data model is cold started and is difficult to operate; and the large amount of data is acquired in the field, so that time and labor are consumed, and the implementation and application of a large data model are limited.
In view of this, a neural network multi-terminal air valve control system and method based on a simulation model are provided.
Disclosure of Invention
Aiming at the problems, the invention provides a neural network multi-terminal air valve control system and method based on a simulation model, a data set is provided for a neural network data model based on an air pipe system simulation model, the problem of insufficient training data of the network data model is solved, and then the integrated control of a multi-terminal variable air volume air conditioning system is realized based on the trained network data model.
In order to achieve the above object, the present invention provides a neural network multi-terminal air valve control system based on a simulation model, comprising: the system comprises a control host, a variable frequency fan and a plurality of variable air volume blast valves;
the control host is connected with the variable-frequency fan through a fan frequency controller, and the variable-frequency fan is connected with the variable-air-volume air valves; the control host is respectively connected with the plurality of variable air volume air valves through a plurality of electric air valve actuators; the plurality of variable air volume air valves are used for conveying wind power and adjusting temperature for a plurality of air-conditioning rooms;
an indoor temperature sensor, an indoor temperature control panel and a temporary air volume sensor are arranged in each of the air-conditioning rooms and are connected to a control host;
the control host is internally stored with an air duct system simulation model and an optimization control program, the optimization control program comprises an air duct model learning module and an air valve opening control module, the air duct system simulation model provides training data for the air duct model learning module, and the air duct model learning module is trained to obtain a neural network data model;
the control host acquires data of the indoor temperature sensor, the indoor temperature control panel, the temporary air quantity sensor, the electric air valve actuator and the fan frequency controller;
the neural network data model calculates the opening degree of each variable air volume air valve according to the acquired data, and transmits the opening degree to the plurality of electric air valve actuators through the air valve opening degree control module to realize the opening degree control of the plurality of variable air volume air valves.
As a further improvement of the invention, the control host comprises an analog signal input interface and an RS485 communication interface;
the indoor temperature sensors, the indoor temperature control panels and the temporary air volume sensors which are arranged in the air-conditioning rooms are all connected with an analog signal input interface;
and the electric air valve actuators and the fan frequency controller are connected with an RS485 communication interface.
As a further improvement of the invention, the analog signal input interface supports 4-20 mA and 0-10V/0-5V physical signals, and the RS485 communication interface supports modbusRTU, modbusSCII and PPI serial port protocols.
As a further improvement of the invention, the optimization control program further comprises a PID control module, and the PID control module controls the frequency of the variable frequency fan through a fan frequency controller.
As a further improvement of the present invention, the neural network data model training process includes:
the air pipe model learning module initializes the parameters of the neural network data model;
the air pipe system simulation model generates a data set which is provided to an air pipe model learning module and used for training and verifying a neural network data model, and the data set comprises the temperature of each air-conditioning room, the opening of each variable air volume air valve, the air volume of each air-conditioning room and the frequency of the variable frequency fan;
and when the difference between the air volume of each air-conditioning room output by the neural network data model and the air volume of each air-conditioning room in the data set under the corresponding working condition is less than 10%, the model training is finished.
The invention also provides a method of the neural network multi-terminal air valve control system based on the simulation model, which comprises the following steps:
and 5, verifying each solution in the feasible solution set through a simulation model of the air pipe system, selecting an optimal solution to be transmitted to each electric air valve actuator, adjusting the opening degree of each variable air volume air valve according to the optimal solution by each electric air valve actuator, and returning to the step 2 for circular execution.
As a further improvement of the present invention,
the system parameters also comprise a variable air volume air valve control period;
and when judging whether the time of the timer is greater than the control period of the variable air volume air valve, if not, increasing the time of the timer by one fan frequency control period, re-acquiring the indoor temperature and indoor temperature set values of each air-conditioning room, and calculating and adjusting the frequency of the variable frequency fan.
As a further improvement of the present invention, each solution in the feasible solution set is verified through a wind pipe system simulation model, and an optimal solution is selected and transmitted to each electric air valve actuator; the method comprises the following steps:
counting the air quantity at each tail end and the required air quantity at each tail end of each solution in the feasible solution set in the air pipe system simulation model;
obtaining a solution which meets the condition whether the sum of the air volume at each tail end is larger than the sum of the air volume required by each tail end and is smaller than 1.5 times of the sum of the air volume required by each tail end;
and taking the solution with the minimum sum of all tail end air volumes as the optimal solution among all solutions meeting the conditions, and transmitting the optimal solution to each electric air valve actuator.
As a further improvement of the present invention, if all solutions in the feasible solution set do not satisfy the condition, a solution with the minimum sum of squared differences between each terminal air volume and each required air volume in all solutions is taken as an optimal solution.
As a further improvement of the invention, the system parameters further include an upper opening limit and a lower opening limit of the variable air volume damper;
if the variable air volume air valve opening is larger than the variable air volume air valve opening upper limit in the optimal solution, taking the variable air volume air valve opening upper limit as the variable air volume air valve opening;
and if the variable air volume air valve opening is smaller than the variable air volume air valve opening lower limit in the optimal solution, taking the variable air volume air valve opening lower limit as the variable air volume air valve opening.
Compared with the prior art, the invention has the beneficial effects that:
according to the invention, training data are provided for the air pipe model learning module through the air pipe system simulation model, so that the problems that the neural network data model is insufficient in training data and part of working conditions are lack of data due to only adopting field data are solved, the neural network data model obtained through training is more accurate, and the opening degree of the multi-terminal variable air volume air valve is more accurate; the problems that in the background art, the opening control of a multi-terminal variable air volume air valve is rough, the precision is low, or PID (proportion integration differentiation) regulation and control are only carried out on a single terminal, the global optimization cannot be achieved, and the requirement on data volume in a big data model is huge are solved.
According to the invention, the frequency of the variable frequency fan is calculated and adjusted according to the current temperature and the temperature set value of each tail end through a PID algorithm, the feasible solution set obtained by the neural network data model is verified one by one through the simulation model again, and the optimal solution is selected, so that the frequency of the variable frequency fan meets the air volume of all air-conditioning rooms, the air volume supply of each tail end can meet the requirement, the air volume waste is reduced, and the energy-saving effect is realized.
Drawings
FIG. 1 is a schematic diagram of a neural network multi-terminal air valve control system based on a simulation model according to an embodiment of the present invention;
fig. 2 is a detailed flowchart of a neural network multi-terminal air valve control method based on a simulation model according to an embodiment of the present invention.
Description of reference numerals:
1. a control host; 2. a CPU operator; 3. a memory; 4. an analog signal input interface; 5. an RS485 communication interface; 6. an electric air valve actuator; 7. a variable air volume air valve; 8. a fan frequency controller; 9. a variable frequency fan; 10. an indoor temperature sensor; 11. an indoor temperature control panel; 12. a temporary air volume sensor; 13. and (4) air-conditioning the room.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, 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 some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
Example (b):
the invention is described in further detail below with reference to the attached drawing figures:
as shown in fig. 1, the present invention provides a neural network multi-terminal air valve control system based on a simulation model, including: the air conditioner comprises a control host 1, a fan frequency controller 8, a variable frequency fan 9, a plurality of electric air valve actuators 6, a plurality of variable air volume air valves 7 and a plurality of air-conditioning rooms 13; the control host 1 comprises a CPU arithmetic unit 2, a memory 3, an analog signal input interface 4 and an RS485 communication interface 5; an indoor temperature sensor 10, an indoor temperature control panel 11 and a temporary air volume sensor 12 are installed in each of the air-conditioning rooms 13 and are connected to the control host 1 through an analog signal input interface 4; the electric air valve actuators 6 are respectively connected with the variable air volume air valves 7 and the RS485 communication interface 5; the input end and the output end of the fan frequency controller 8 are respectively connected with the RS485 communication interface 5 and the variable frequency fan 9, so that the rotating speed of the variable frequency fan 9 is controlled; the variable-frequency fan 9 is connected with the variable-air-volume air valves 7, and the variable-air-volume air valves 7 respectively control the air volume of the air-conditioned rooms 13 and regulate the indoor temperature;
the control host 1 stores an air duct system simulation model and an optimization control program, the optimization control program comprises an air duct model learning module, an air valve opening control module and a PID control module, the air duct system simulation model provides training data for the air duct model learning module, and the air duct model learning module obtains a neural network data model through training;
the control host 1 acquires data of the indoor temperature sensor 10, the indoor temperature control panel 11, the temporary air quantity sensor 12, the electric air valve actuator 6 and the fan frequency controller 8; the PID control module calculates and adjusts the frequency of the variable frequency fan 9 according to the data of the indoor temperature sensor 10 and the indoor temperature control panel 11; the neural network data model calculates the opening degree of each variable air volume air valve 7 according to the acquired data, and transmits the opening degree to the plurality of electric air valve actuators 6 through the air valve opening degree control module, so that the opening degree control of the plurality of variable air volume air valves 7 is realized.
The analog signal input interface 4 supports 4-20 mA and 0-10V/0-5V physical signals, and the RS485 communication interface supports modbusRTU, modbusSCII and PPI serial port protocols.
The training process of the neural network data model comprises the following steps:
(1) establishing a fan system simulation model in advance, and correcting the simulation model according to the temporary air volume sensors 12 of the air-conditioning rooms 13;
(2) the air pipe model learning module initializes parameters of a neural network data model and mainly comprises the number of input layers, the number of hidden layers, the number of output layers and an activation function of the neural network model;
(3) generating a data set based on the corrected fan system simulation model, and providing the data set to an air pipe model learning module; the data set comprises the temperature of each air-conditioning room 13, the set temperature of each air-conditioning room 13, the opening degree of each variable air volume air valve 7, the air volume of each air-conditioning room 13 and the frequency of the variable frequency fan 9;
(4) the air duct model learning module uses 85% of data in the data set as a training set, and 15% of data is used for verification;
(5) and (3) verifying the neural network data model:
inputting 15% of the data in the data set for validation into a neural network data model, the input data comprising: the temperature of each air-conditioning room 13, the set temperature of each air-conditioning room 13 and the frequency of the variable-frequency fan 9, the opening degree of the variable air volume air valve 7 corresponding to the output of the neural network data model, the opening degree of the variable air volume air valve 7 is input into the simulation model to calculate the air volume of each air-conditioning room, and the calculated air volume of each air-conditioning room 13 is compared with the air volume of each air-conditioning room 13 corresponding to the group of data in the data set to obtain an error;
and if the difference between the air volume of each air-conditioning room 13 output by the neural network data model and the air volume of each air-conditioning room 13 in the data set under the corresponding working condition is less than 10%, the model training is considered to be finished. Otherwise, modifying the model, and re-tuning the model parameters until the difference between the air volume signal data collected by the temporary air volume sensor 12 and the air volume data output by the air duct system simulation model in the memory 3 is less than 10%.
As shown in fig. 2, the present invention further provides a method of a neural network multi-terminal air valve control system based on a simulation model, which includes the steps of:
s1, initializing system parameters, where the system parameters include PID control module parameters, specifically, a P value, an I value, a D value, a frequency control period Δ T of the variable frequency fan 9, a control period Δ T of the variable air volume damper 7, an opening upper limit Kmax of the variable air volume damper 7, and an opening lower limit Kmin of the variable air volume damper 7, and setting a timer time T to 0, that is, T is 0;
s2, the control host 1 respectively obtains indoor temperature and indoor temperature set values through the indoor temperature sensors 10 and the indoor temperature control panels 11 of the air-conditioning rooms 13, and transmits the set values into an optimized control program in the memory 3 through the analog signal input interface 4;
s3, the PID control module calculates and adjusts the frequency of the variable frequency fan 9 by utilizing a PID algorithm according to the indoor temperature and the indoor temperature set value of each air-conditioning room 13;
s4, judging whether the timer time t is larger than delta t by the system;
if yes, the process goes to S5 when the refresh timer reaches time t equal to 0;
if the judgment result is negative, increasing the time of the timer by a control period delta T of the variable air volume air valve 7, namely T is T plus delta T, and returning to S3 for cyclic execution until T is greater than delta T;
s5, acquiring the frequency of the current variable frequency fan 9 and the indoor temperature and indoor temperature set value of each air-conditioning room 13 through an RS485 protocol, and inputting the frequency, the indoor temperature and the indoor temperature set value into a neural network data model, wherein the neural network data model specifically comprises the following steps:
(1) the opening degree signal of the variable air volume air valve 7 obtained by the electric air valve actuator 6 is transmitted into a neural network data model in an optimization control program in the memory 3 through the RS485 communication interface 5;
(2) the real-time frequency signal of the variable frequency fan 9 is obtained by the fan frequency controller 8 and is transmitted into a neural network data model in an optimization control program in the memory 3 through the RS485 communication interface 5,
(3) the temperature signal data of the indoor temperature set values of the indoor temperature sensors 10 and the indoor temperature control panel 11 are transmitted to the neural network data model in the optimization control program through the analog signal input interface 4.
S6, the CPU arithmetic unit 2 runs the neural network data model and outputs the opening feasible solution set R (n) of each variable air volume blast gate 7;
the feasible solution set R (n) comprises a plurality of groups of solutions, wherein the feasible solutions are a certain number of feasible solutions obtained by running the neural network model for a certain number of times, and each group of solutions can meet the requirement of each tail end air volume.
S7, verifying each solution R in the feasible solution set R (n) through an air pipe system simulation model, and counting the air quantity Qsupply (i) of each air-conditioning room 13 and the required air quantity Qneed (i) of each air-conditioning room 13, which are obtained by each feasible solution R in the air pipe system simulation model;
s8, judging whether Qsupplied (i) > Qneed (i) exists in the feasible solution set R (n), if yes, taking the solution with the minimum value of the sigma Qsupplied (i) as the opening value of each variable air volume air valve 7, and entering S9;
if not, taking the solution with the minimum sum of squared differences between the air volume qsupplied (i) of each air-conditioning room 13 and the air volume required qneed (i) as the optimal solution, namely the solution with the minimum [ Σ (qsupplied (i) -qned (i))2] as the opening value of each variable air volume damper 7, and entering S9;
s9, judging whether the opening value of each corresponding variable air volume air valve 7 in the optimal solution is larger than the upper opening limit Kmax of the variable air volume air valve 7, if so, taking the upper opening limit Kmax of the variable air volume air valve 7 as the opening value of the variable air volume air valve 7; otherwise, further judging whether the corresponding opening degree value of the variable air volume air valve 7 in the optimal solution is smaller than the lower opening degree limit Kmin of the variable air volume air valve 7, and taking the lower opening degree limit Kmin of the variable air volume air valve 7 as the opening degree of the variable air volume air valve 7;
and S10, sending the new opening values of the variable air volume blast valves 7 into the electric blast valve actuators 6 through the RS485 communication interface 5, realizing the opening control of the variable air volume blast valves 7, and simultaneously returning to S2 to continue the circular execution.
The invention has the advantages that:
(1) training data are provided for the air pipe model learning module through the air pipe system simulation model, the problems that the neural network data model is insufficient in training data and part of working conditions lack data due to the fact that only field data are adopted are solved, the neural network data model obtained through training is more accurate, and the opening degree of the multi-terminal variable air volume air valve is more accurate; the problems that in the background art, the opening control of a multi-terminal variable air volume air valve is rough, the precision is low, or PID (proportion integration differentiation) regulation and control are only carried out on a single terminal, the global optimization cannot be achieved, and the requirement on data volume in a big data model is huge are solved.
(2) And calculating and adjusting the frequency of the variable frequency fan according to the current temperature and the temperature set value of each terminal by a PID algorithm, verifying a feasible solution set obtained by the neural network data model one by a simulation model again, and selecting an optimal solution, so that the frequency of the variable frequency fan meets the air volume of all air-conditioning rooms, the air volume supply of each terminal can meet the requirement, the air volume waste is reduced, and the energy-saving effect is realized.
The above is only a preferred embodiment of the present invention, and is not intended to limit the present invention, and various modifications and changes will occur to those skilled in the art. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (10)
1. A neural network multi-terminal air valve control system based on a simulation model is characterized by comprising the following components: the system comprises a control host, a variable frequency fan and a plurality of variable air volume blast valves;
the control host is connected with the variable-frequency fan through a fan frequency controller, and the variable-frequency fan is connected with the variable-air-volume air valves; the control host is respectively connected with the plurality of variable air volume air valves through a plurality of electric air valve actuators; the plurality of variable air volume air valves are used for conveying wind power and adjusting temperature for a plurality of air-conditioning rooms;
an indoor temperature sensor, an indoor temperature control panel and a temporary air volume sensor are arranged in each of the air-conditioning rooms and are connected to a control host;
the control host is internally stored with an air duct system simulation model and an optimization control program, the optimization control program comprises an air duct model learning module and an air valve opening control module, the air duct system simulation model provides training data for the air duct model learning module, and the air duct model learning module is trained to obtain a neural network data model;
the control host acquires data of the indoor temperature sensor, the indoor temperature control panel, the temporary air quantity sensor, the electric air valve actuator and the fan frequency controller;
the neural network data model calculates the opening degree of each variable air volume air valve according to the acquired data, and transmits the opening degree to the plurality of electric air valve actuators through the air valve opening degree control module to realize the opening degree control of the plurality of variable air volume air valves.
2. The control system of claim 1, wherein: the control host comprises an analog signal input interface and an RS485 communication interface;
the indoor temperature sensors, the indoor temperature control panels and the temporary air volume sensors which are arranged in the air-conditioning rooms are all connected with an analog signal input interface;
and the electric air valve actuators and the fan frequency controller are connected with an RS485 communication interface.
3. The control system of claim 1, wherein the analog signal input interface supports 4-20 mA and 0-10V/0-5V physical signals, and the RS485 communication interface supports modbusRTU, modbusscii, PPI serial protocol.
4. The control system of claim 1, wherein: the optimization control program further comprises a PID control module, and the PID control module controls the frequency of the variable-frequency fan through a fan frequency controller.
5. The control system of claim 1, wherein: the neural network data model training process comprises:
the air pipe model learning module initializes the parameters of the neural network data model;
the air pipe system simulation model generates a data set which is provided to an air pipe model learning module and used for training and verifying a neural network data model, and the data set comprises the temperature of each air-conditioning room, the opening of each variable air volume air valve, the air volume of each air-conditioning room and the frequency of the variable frequency fan;
and when the difference between the air volume of each air-conditioning room output by the neural network data model and the air volume of each air-conditioning room in the data set under the corresponding working condition is less than 10%, the model training is finished.
6. A method based on the control system of claims 1-5, characterized by comprising the steps of:
step 1, initializing system parameters, wherein the system parameters comprise PID control module parameters and a fan frequency control period, and setting the time of a timer to be 0;
step 2, the control host acquires the indoor temperature and indoor temperature set value of each air-conditioning room;
step 3, the PID control module calculates and adjusts the frequency of the variable frequency fan according to the indoor temperature and the indoor temperature set value of each air-conditioning room, and judges whether the time of the timer is greater than the control period of the variable air volume air valve;
step 4, if the judgment result is yes, resetting the time of the timer to be 0, acquiring the frequency of the current variable frequency fan and the variable air volume air valve opening, the indoor temperature and the indoor temperature set value of each air-conditioning room, inputting the frequency, the variable air volume air valve opening, the indoor temperature and the indoor temperature set value into a neural network data model, and calculating a feasible solution set of each variable air volume air valve opening;
and 5, verifying each solution in the feasible solution set through a simulation model of the air pipe system, selecting an optimal solution to be transmitted to each electric air valve actuator, adjusting the opening degree of each variable air volume air valve according to the optimal solution by each electric air valve actuator, and returning to the step 2 for circular execution.
7. The method of claim 6, wherein:
the system parameters also comprise a variable air volume air valve control period;
and when judging whether the time of the timer is greater than the control period of the variable air volume air valve, if not, increasing the time of the timer by one fan frequency control period, re-acquiring the indoor temperature and indoor temperature set values of each air-conditioning room, and calculating and adjusting the frequency of the variable frequency fan.
8. The method of claim 6, wherein: verifying each solution in the feasible solution set through a wind pipe system simulation model, and selecting an optimal solution to transmit to each electric air valve actuator; the method comprises the following steps:
counting the air quantity at each tail end and the required air quantity at each tail end of each solution in the feasible solution set in the air pipe system simulation model;
obtaining a solution which meets the condition whether the sum of the air volume at each tail end is larger than the sum of the air volume required by each tail end and is smaller than 1.5 times of the sum of the air volume required by each tail end;
and taking the solution with the minimum sum of all tail end air volumes as the optimal solution among all solutions meeting the conditions, and transmitting the optimal solution to each electric air valve actuator.
9. The method of claim 8, wherein: and if all solutions in the feasible solution set do not meet the conditions, taking the solution with the minimum sum of square differences of the terminal air volume and the required air volume in all solutions as the optimal solution.
10. The method of claim 6, wherein: the system parameters also comprise an upper opening limit of the variable air volume air valve and a lower opening limit of the variable air volume air valve;
if the variable air volume air valve opening is larger than the variable air volume air valve opening upper limit in the optimal solution, taking the variable air volume air valve opening upper limit as the variable air volume air valve opening;
and if the variable air volume air valve opening is smaller than the variable air volume air valve opening lower limit in the optimal solution, taking the variable air volume air valve opening lower limit as the variable air volume air valve opening.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202120953508 | 2021-05-07 | ||
| CN2021209535089 | 2021-05-07 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113418288A true CN113418288A (en) | 2021-09-21 |
Family
ID=77720017
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110746780.4A Pending CN113418288A (en) | 2021-05-07 | 2021-07-01 | Simulation model-based neural network multi-terminal air valve control system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113418288A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114216256A (en) * | 2021-12-22 | 2022-03-22 | 中国海洋大学 | Ventilation system air volume control method of off-line pre-training-on-line learning |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109990445A (en) * | 2019-04-09 | 2019-07-09 | 上海真聂思楼宇科技有限公司 | A kind of air conditioning system with variable energy-saving controller and method |
| CN110736228A (en) * | 2019-10-29 | 2020-01-31 | 珠海格力电器股份有限公司 | Control method and device of air conditioning equipment, storage medium and air conditioning system |
| CN110736244A (en) * | 2019-10-29 | 2020-01-31 | 珠海格力电器股份有限公司 | Method and device for controlling air conditioner and air conditioner |
| CN110848896A (en) * | 2019-11-12 | 2020-02-28 | 深圳孚沃德斯科技有限公司 | Intelligent energy-saving control system and method for air conditioner cooling system based on neural network |
| CN111288610A (en) * | 2020-02-13 | 2020-06-16 | 西安建筑科技大学 | Variable static pressure self-adaptive fuzzy control method for variable air volume air conditioning system |
| CN112648717A (en) * | 2020-12-16 | 2021-04-13 | 桂林电子科技大学 | Terminal control method of multi-zone variable air volume air conditioning system based on radial basis function neural network |
-
2021
- 2021-07-01 CN CN202110746780.4A patent/CN113418288A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109990445A (en) * | 2019-04-09 | 2019-07-09 | 上海真聂思楼宇科技有限公司 | A kind of air conditioning system with variable energy-saving controller and method |
| CN110736228A (en) * | 2019-10-29 | 2020-01-31 | 珠海格力电器股份有限公司 | Control method and device of air conditioning equipment, storage medium and air conditioning system |
| CN110736244A (en) * | 2019-10-29 | 2020-01-31 | 珠海格力电器股份有限公司 | Method and device for controlling air conditioner and air conditioner |
| CN110848896A (en) * | 2019-11-12 | 2020-02-28 | 深圳孚沃德斯科技有限公司 | Intelligent energy-saving control system and method for air conditioner cooling system based on neural network |
| CN111288610A (en) * | 2020-02-13 | 2020-06-16 | 西安建筑科技大学 | Variable static pressure self-adaptive fuzzy control method for variable air volume air conditioning system |
| CN112648717A (en) * | 2020-12-16 | 2021-04-13 | 桂林电子科技大学 | Terminal control method of multi-zone variable air volume air conditioning system based on radial basis function neural network |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114216256A (en) * | 2021-12-22 | 2022-03-22 | 中国海洋大学 | Ventilation system air volume control method of off-line pre-training-on-line learning |
| CN114216256B (en) * | 2021-12-22 | 2022-09-23 | 中国海洋大学 | Ventilation system air volume control method of off-line pre-training-on-line learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Okochi et al. | A review of recent developments and technological advancements of variable-air-volume (VAV) air-conditioning systems | |
| JP5575784B2 (en) | System and method for controlling energy consumption efficiency | |
| CN109990445B (en) | Energy-saving controller and method for variable air volume air conditioning system | |
| CN104019526B (en) | Improve PSO algorithm Fuzzy Adaptive PID temperature and humidity control system and method | |
| CN109798646B (en) | Variable air volume air conditioner control system and method based on big data platform | |
| CN109612045A (en) | A kind of gravity-flow ventilation suitable for building conditioning in Transition Season and force ventilation converting system | |
| CN105047058A (en) | Full-scale central air-conditioning and centralized heating integration experiment platform and control method therefor | |
| CN114265306B (en) | Online wind balance control method and control system for ocean platform ventilation system | |
| CN104390323A (en) | Electric temperature difference control valve of air-conditioning system and temperature difference control method of electric temperature difference control valve | |
| CN101782261A (en) | Nonlinear self-adapting energy-saving control method for heating ventilation air-conditioning system | |
| CN107940666B (en) | Real-time automatic adjusting and controlling device for classroom fresh air purifying system | |
| CN204946458U (en) | A kind of full size central air-conditioning and central heating synthesis experiment platform | |
| WO2022062339A1 (en) | System and method for controlling air valve of variable air volume cabin unit | |
| CN108759064A (en) | A kind of method for noise reduction control, device and multi-online air-conditioning system | |
| CN111047117B (en) | Differential pressure bypass valve energy-saving optimization method based on terminal load prediction | |
| CN203396032U (en) | Room temperature control device based on fuzzy self-adaption PID (proportion integration differentiation) | |
| CN109612047A (en) | The supply air temperature control method of air conditioning system with variable | |
| CN113418288A (en) | Simulation model-based neural network multi-terminal air valve control system and method | |
| CN101922779A (en) | Fan coil system and control method and device thereof | |
| CN111444627A (en) | Comfort zone energy-saving optimization method based on indoor quality control model | |
| Zhao et al. | An optimal control method for discrete variable outdoor air volume setpoint determination in variable air volume systems | |
| CN112648717B (en) | Terminal control method of multi-zone variable air volume air conditioning system based on radial basis function neural network | |
| CN114110940A (en) | Intelligent electric regulating valve adjusting method and system for air-conditioning water system | |
| CN1837696A (en) | Control method for air quantity-changeable air conditioning system employing total air quantity calculation method | |
| CN106440154B (en) | Air conditioning system control method with humidifying function and air conditioning 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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210921 |
|
| RJ01 | Rejection of invention patent application after publication |