CN108895618B - Air conditioner control method based on neural network - Google Patents
Air conditioner control method based on neural network Download PDFInfo
- Publication number
- CN108895618B CN108895618B CN201811155298.8A CN201811155298A CN108895618B CN 108895618 B CN108895618 B CN 108895618B CN 201811155298 A CN201811155298 A CN 201811155298A CN 108895618 B CN108895618 B CN 108895618B
- Authority
- CN
- China
- Prior art keywords
- control method
- neural network
- conditioning system
- control
- air 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
- F24F11/50—Control or safety arrangements characterised by user interfaces or communication
- F24F11/56—Remote control
- F24F11/58—Remote control using Internet communication
-
- 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
- F24F11/30—Control or safety arrangements for purposes related to the operation of the system, e.g. for safety or monitoring
- F24F11/46—Improving electric energy efficiency or saving
- F24F11/47—Responding to energy costs
-
- 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
-
- 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
- F24F11/80—Control systems characterised by their outputs; Constructional details thereof for controlling the temperature of the supplied air
- F24F11/83—Control systems characterised by their outputs; Constructional details thereof for controlling the temperature of the supplied air by controlling the supply of heat-exchange fluids to heat-exchangers
- F24F11/84—Control systems characterised by their outputs; Constructional details thereof for controlling the temperature of the supplied air by controlling the supply of heat-exchange fluids to heat-exchangers using valves
-
- 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
- F24F11/80—Control systems characterised by their outputs; Constructional details thereof for controlling the temperature of the supplied air
- F24F11/86—Control systems characterised by their outputs; Constructional details thereof for controlling the temperature of the supplied air by controlling compressors within refrigeration or heat pump circuits
Landscapes
- Engineering & Computer Science (AREA)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Human Computer Interaction (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Thermal Sciences (AREA)
- Fluid Mechanics (AREA)
- Air Conditioning Control Device (AREA)
Abstract
The invention relates to an air conditioner control method based on a neural network, which combines the neural network control method with the traditional control method, automatically detects the observed quantity of an air conditioning system, and applies the generated controlled quantity to the air conditioning system for automatic control, so that the refrigerating capacity, the heating capacity or the energy consumption of the air conditioning system are always kept at expected values; wherein, the traditional control method comprises the following steps: a fuzzy control method or a PID control method. The invention ensures that the air conditioning system has higher refrigerating capacity or heating capacity and less energy consumption under the condition of finally achieving the same control target.
Description
Technical Field
The invention relates to an air conditioner control method, in particular to an air conditioner control method based on a neural network.
Background
Air conditioning is the maintenance of ambient state parameters in an indoor space (e.g. a building, a train, an airplane, etc.) at desired values throughout varying outdoor climatic conditions and indoor loads. The air conditioner automatic control is that the control parameters (control quantity) of the air conditioner equipment are automatically adjusted through automatic detection of the air state parameters (observed quantity), so that the air conditioner system is kept in an optimal working state, and the refrigerating capacity, the heating capacity or the energy consumption of the air conditioner system is always kept at an expected value.
The main air state parameters (observed quantities) are: indoor and outdoor ambient temperatures, indoor and outdoor ambient humidity, surface temperatures of condensers and evaporators, and the like. The main control parameters (control amounts) are: compressor frequency, indoor unit fan speed, outdoor unit fan speed, expansion valve opening, and the like.
In the prior art, a fuzzy control method and a PID (proportional integral derivative) control method are mainly used to adjust a controlled variable based on a measurement result of an observed variable, thereby realizing automatic control of an air conditioner.
The fuzzy control method is essentially a nonlinear control, and belongs to the field of intelligent control. The fuzzy control method has the great characteristics of not only having a systematized theory but also having a large number of practical application backgrounds. The method mainly comprises the steps of variable definition, fuzzification, knowledge base, logic judgment, defuzzification and the like.
The PID control method is a closed-loop feedback control method which is most widely applied in engineering practice at present and is formed by combining a proportional unit P, an integral unit I and a differential unit D. The basis of the PID control method is proportional control, the output of which is proportional to the input error signal; the output of the integral control is in a direct proportion relation with the integral of the input error signal, so that the steady-state error can be eliminated, but the overshoot can be increased; the output of the differential control is in direct proportion to the differential of the input error signal (namely the change rate of the error), so that the response speed of the large inertia system can be accelerated, and the overshoot trend can be weakened.
However, both of the above-mentioned conventional control methods have the following problems: 1. because of being limited by the capacity of a processor (mainly a single chip microcomputer), algorithm designs of a fuzzy control method and a PID control method are simplified; 2. the fuzzy control method and the PID control method are both carried out based on limited data samples, and meanwhile, the control parameters of manual optimization are relied on, so that the method cannot be optimally applied to all working scenes; 3. when high dimensional problems are involved, i.e. problems with more than one observed quantity resulting in more than one controlled quantity, the tuning optimization of the fuzzy control method and the PID control method becomes extremely complex. The conventional control method has the problem that the air conditioning system has low energy efficiency in cooling or heating.
The neural network is a network system with intelligent information processing functions of learning, association, memory, pattern recognition and the like by modeling and connecting the basic unit of the human brain, namely neurons and exploring a model for simulating the functions of the human brain nervous system. The neural network has the characteristics of large-scale parallel processing, distributed information storage, good self-organizing and self-learning capabilities and the like.
The neural network is formed by connecting a plurality of 'neuron' units, and the basic structure of each neuron consists of a linear transformation unit with adjustable weight and a nonlinear output function (activation function). The multilayer neural network has strong nonlinear mapping capability, can approach any (linear or nonlinear) function theoretically, and has great flexibility because parameters such as the number of middle layers of the network, the number of processing units of each layer, the learning coefficient of the network and the like can be set according to specific conditions.
The bp (back propagation) algorithm, also called an error back propagation algorithm, is a commonly used weight updating algorithm in an artificial neural network, or called a learning algorithm. The neural network can automatically adjust network parameters (weight) according to data by using a BP algorithm, so that the neural network can automatically adapt to the environment and summarize rules, and has wide application prospects in many fields such as optimization, signal processing and mode recognition, intelligent control, fault diagnosis and the like.
Based on the above, the present invention provides an air conditioner control method based on a neural network, so as to solve the disadvantages and limitations in the prior art.
Disclosure of Invention
The invention aims to provide an air conditioner control method based on a neural network, which enables an air conditioner system to have higher cooling capacity or heating capacity and less energy consumption under the condition of ensuring that the same control target is finally achieved.
In order to achieve the aim, the invention provides an air conditioner control method based on a neural network, which combines the neural network control method with the traditional control method, automatically detects the observed quantity of an air conditioning system, and applies the generated control quantity to the air conditioning system for automatic control, so that the refrigerating capacity, the heating capacity or the energy consumption of the air conditioning system are always kept at expected values; wherein, the traditional control method comprises the following steps: a fuzzy control method or a PID control method.
The method for keeping the refrigerating capacity, the heating capacity or the energy consumption of the air conditioning system at a desired value all the time specifically comprises the following steps: when the absolute value of the difference between the refrigerating capacity or the heating capacity of the air conditioning system and the refrigerating capacity or the heating capacity obtained by the traditional control method is smaller than a certain threshold value, the energy consumption of the air conditioning system is kept to be not higher than that of the traditional control method; and when the absolute value of the difference between the energy consumption of the air conditioning system and the energy consumption required by the traditional control method is smaller than a certain threshold, keeping the refrigerating capacity or the heating capacity of the air conditioning system not lower than that of the traditional control method.
In a preferred embodiment of the present invention, the neural network-based air conditioner control method includes the following steps:
SA1, adopting traditional control method, fuzzy control method or PID control method, to automatically detect the observed quantity of air-conditioning system, and outputting corresponding state quantity;
SA2, controlling and regulating the state quantity by adopting a neural network control method, and outputting a corresponding control quantity;
SA3, and automatically controlling the control amount by applying the control amount to the air conditioning system.
The state quantity comprises at least one or any combination of the following variables: the control amount generated by the conventional control method; estimating the required refrigerating capacity or other parameters corresponding to the required refrigerating capacity based on the observed quantity; estimating other parameters based on the observed quantity; measuring the observed quantity; the amount of control fed back.
The state quantity comprises a current state quantity and/or a historical state quantity.
In a preferred embodiment of the present invention, the neural network-based air conditioner control method includes the following steps:
SB1, adopting a neural network control method to control and regulate the observed quantity and/or the feedback control quantity of the air conditioning system, and outputting a new observed quantity;
SB2, adopting traditional control method, namely fuzzy control method or PID control method, to automatically detect the new observed quantity and output corresponding control quantity;
SB3, the control amount is applied to the air conditioning system to automatically control the air conditioning system.
In a preferred embodiment of the present invention, the neural network-based air conditioner control method includes the following steps:
SC1, fitting the traditional control method, namely fuzzy control method or PID control method, into the neural network control method with the same or similar refrigerating and/or heating capability and energy consumption;
SC2, adopting the neural network control method fitting the traditional control method, automatically detecting and controlling and adjusting the observed quantity and/or feedback control quantity of the air-conditioning system, and outputting the corresponding control quantity;
and SC3, applying the control quantity to the air conditioning system to automatically control.
The refrigerant with similar refrigerating and/or heating capacity and energy consumption specifically refers to: when the absolute value of the difference between the refrigerating capacity or the heating capacity of the air conditioning system obtained by adopting the neural network control method fitted by the traditional control method and the refrigerating capacity or the heating capacity obtained by adopting the traditional control method is smaller than a certain threshold value, the energy consumption of the air conditioning system is kept to be not higher than that of the traditional control method; and when the absolute value of the difference between the energy consumption of the air conditioning system obtained by adopting the neural network control method fitted by the traditional control method and the energy consumption required by the traditional control method is smaller than a certain threshold value, keeping the refrigerating capacity or the heating capacity of the air conditioning system not lower than that of the traditional control method.
Wherein the control amount may be further subjected to differential control.
Wherein, the control quantity comprises at least one or any combination of the following variables: compressor frequency, indoor unit fan speed, outdoor unit fan speed, expansion valve opening.
In a preferred embodiment of the present invention, the neural network adopts a cascade structure, and is formed by cascading at least one basic network and at least one control quantity generation network; the basic network carries out feature extraction on the input state quantity, the observed quantity and the feedback control quantity of the air conditioning system to obtain the intermediate quantity of the neural network, and the control quantity of the air conditioning system is output through the network generated by the control quantity.
In a preferred embodiment of the present invention, the neural network adopts a hybrid structure and is formed by cascading a state classification network and a plurality of control quantity generation networks; the state classification network classifies the input state quantity, observation quantity and feedback control quantity of the air conditioning system; and each control quantity generation network correspondingly outputs the control quantity of the air conditioning system according to the classification result of each type.
In summary, compared with the conventional control method, the neural network-based air conditioner control method provided by the invention has a better control effect under the condition of ensuring that the same control target is finally achieved, so that the air conditioner system has higher refrigerating capacity or heating capacity and less energy consumption.
Drawings
Fig. 1 is a schematic process diagram of a first embodiment of an air conditioner control method based on a neural network according to the present invention;
fig. 2 is a process diagram illustrating a second embodiment of the neural network-based air conditioner control method according to the present invention;
FIG. 3 is a process diagram illustrating a third embodiment of the neural network-based air conditioning control method according to the present invention;
FIG. 4 is a schematic diagram of a neural network control method according to the present invention in comparison with a conventional control method;
FIG. 5 is a schematic diagram of a neural network in a cascade configuration according to the present invention;
FIG. 6 is a schematic diagram of a neural network employing a hybrid architecture in accordance with the present invention;
fig. 7 is a process diagram illustrating a specific example of an air conditioner control method based on a neural network according to the present invention;
fig. 8 is a schematic process diagram of the neural network-based air conditioner control method in the present invention, which employs differential control.
Detailed Description
The technical contents, construction features, achieved objects and effects of the present invention will be described in detail by preferred embodiments with reference to fig. 1 to 8.
The air conditioner control method based on the neural network is different from the traditional control method, the control quantity directly generated by a fuzzy control method or a PID control method is acted on the air conditioner system for automatic control, but the neural network control method is combined with the traditional control method, the observation quantity of the air conditioner system is automatically detected, and the generated control quantity is acted on the air conditioner system for automatic control, so that the refrigerating capacity, the heating capacity or the energy consumption of the air conditioner system always keeps an expected value; wherein, the traditional control method comprises the following steps: a fuzzy control method or a PID control method.
The method for keeping the refrigerating capacity, the heating capacity or the energy consumption of the air conditioning system at a desired value all the time specifically includes: the air conditioning system always maintains the same or similar cooling capacity, heating capacity or energy consumption as the traditional control method. Similar cooling or heating amounts referred to herein are those slightly higher than conventional control methods; and similar energy consumption means slightly lower energy consumption than that of the conventional control method.
In other words, the keeping of the cooling capacity, the heating capacity, or the energy consumption of the air conditioning system at a desired value is specifically as follows: when the absolute value of the difference between the refrigerating capacity or the heating capacity of the air-conditioning system and the refrigerating capacity or the heating capacity obtained by the traditional control method is smaller than a certain threshold value, the energy consumption of the air-conditioning system is kept to be not higher than that of the traditional control method with a larger probability; when the absolute value of the difference between the energy consumption of the air conditioning system and the energy consumption required by the traditional control method is smaller than a certain threshold value, the refrigerating capacity or the heating capacity of the air conditioning system is kept to be not lower than that of the traditional control method with a larger probability.
As shown in fig. 1, a first embodiment of the neural network-based air conditioner control method according to the present invention includes the following steps:
SA1, adopting traditional control method, fuzzy control method or PID control method, to automatically detect the observed quantity of air-conditioning system, and outputting corresponding state quantity;
SA2, controlling and regulating the state quantity by adopting a neural network control method, and outputting a corresponding control quantity;
SA3, and automatically controlling the control amount by applying the control amount to the air conditioning system.
The state quantity comprises at least one or any combination of the following variables: the control amount generated by the conventional control method; estimating the required refrigerating capacity or other parameters corresponding to the required refrigerating capacity based on the observed quantity; estimating other parameters based on the observed quantity; the observed quantity itself; the amount of control fed back is itself.
The state quantity comprises a current state quantity and/or a historical state quantity.
In a preferred embodiment of the present invention, as shown in fig. 8, the above-mentioned neural network-based air conditioner control method may be further optimized. Carrying out differential control on the output of the neural network, specifically: the output of the neural network is taken as a relative control quantity, namely a time difference quantity of the control quantity, and the relative control quantity is further added with the control quantity obtained at the previous moment to obtain a real control quantity which is applied to the air conditioning system, and meanwhile, the currently obtained control quantity also needs to participate in the calculation of the real control quantity at the next moment.
As shown in fig. 2, a second embodiment of the neural network-based air conditioner control method according to the present invention, which is equivalent to the method according to the first embodiment, includes the following steps:
SB1, adopting a neural network control method to control and regulate the observed quantity and/or the feedback control quantity of the air conditioning system, and outputting a new observed quantity;
SB2, adopting traditional control method, namely fuzzy control method or PID control method, to automatically detect the new observed quantity and output corresponding control quantity;
SB3, the control amount is applied to the air conditioning system to automatically control the air conditioning system.
In a preferred embodiment of the present invention, the differential control may also be performed on the control quantity, which is the same as the method described in the first embodiment, and is not described herein again.
As shown in fig. 3, a third embodiment of the neural network-based air conditioner control method according to the present invention, which is equivalent to the method according to the first embodiment, includes the following steps:
SC1, fitting the traditional control method, namely fuzzy control method or PID control method, into the neural network control method with the same or similar refrigerating and/or heating capability and energy consumption;
SC2, adopting the neural network control method fitting the traditional control method, automatically detecting and controlling and adjusting the observed quantity and/or feedback control quantity of the air-conditioning system, and outputting the corresponding control quantity;
and SC3, applying the control quantity to the air conditioning system to automatically control.
The refrigerant with similar refrigerating and/or heating capacity specifically refers to: the refrigerating capacity or the heating capacity of the air conditioning system is slightly higher than that of the traditional control method; the method has similar energy consumption, and specifically comprises the following steps: the energy consumption of the air conditioning system is slightly lower than that of the traditional control method.
That is, when the absolute value of the difference between the cooling capacity or heating capacity of the air conditioning system obtained by the neural network control method fitted by the conventional control method and the cooling capacity or heating capacity obtained by the conventional control method is less than a certain threshold, the energy consumption of the air conditioning system is kept to be not higher than that of the conventional control method with a relatively high probability; when the absolute value of the difference between the energy consumption of the air conditioning system obtained by adopting the neural network control method fitted by the traditional control method and the energy consumption required by the traditional control method is smaller than a certain threshold value, the refrigerating capacity or the heating capacity of the air conditioning system is kept to be not lower than that of the traditional control method with a larger probability.
In the present embodiment, a separate neural network control method is adopted, and the separate neural network is used to express the control logic of the conventional control method in addition to the control logic of the neural network in the first embodiment. That is, in the present embodiment, the control logic and the corresponding characteristics of capability, energy consumption, etc. (which may be expressed as a complex nonlinear mapping relationship) of the conventional control method are fitted to be expressed by using the neural network, so that the related steps of the conventional control method are not required.
In a preferred embodiment of the present invention, the differential control may also be performed on the control quantity, which is the same as the method described in the first embodiment, and is not described herein again.
For the three embodiments, the control quantity includes at least one or any combination of the following variables: compressor frequency, indoor unit fan speed, outdoor unit fan speed, expansion valve opening.
In the invention, the neural network consists of a plurality of groups of neurons connected in parallel or in series, each neuron corresponds to an independent nonlinear mapping function Y = f (x), and the whole neural network expresses a more complex mapping function Y = f (x).
In a preferred embodiment of the present invention, the neural network adopts a cascade structure, and is formed by cascading at least one basic network (feature extraction network) and at least one control quantity generation network; the basic network carries out feature extraction on the input state quantity, the observed quantity and the feedback control quantity of the air conditioning system to obtain the intermediate quantity of the neural network, and the control quantity of the air conditioning system is output through the network generated by the control quantity.
In actual use, the network of at least one stage can be dynamically updated and configured according to the statistical data of the observed quantity of the air conditioning system. For example, neural networks typically have very large parameters that define the functional expression F for each neuron and the functional expression F for the entire neural network. In this embodiment, the parameters of the control quantity generation network or the basic network may be dynamically configured and updated according to information such as different regions, seasons, or user habits, that is, the mapping function F of a plurality of neurons or the mapping function F of the entire neural network may be dynamically configured and updated.
As shown in fig. 5, a typical neural network with two stages connected in series is shown. The intermediate quantities generated by the base network may not generally have actual physical significance, collectively referred to as features, so the base network may also be referred to as a feature extraction network. The control quantity generation network cascaded with the basic network is used for generating output with real physical meaning, namely output control quantity.
In a preferred embodiment of the present invention, the neural network adopts a hybrid structure, and is formed by cascading a state classification network and a plurality of control quantity generation networks, as shown in fig. 6; the state classification network classifies the input state quantity, observation quantity and feedback control quantity of the air conditioning system, and comprises classification algorithm based on neural network, or clustering algorithm, or fuzzy logic algorithm; and each control quantity generation network correspondingly outputs the control quantity of the air conditioning system according to the classification result of each type.
In actual use, at least one of the control quantity generation networks can be dynamically updated and configured according to the classification result of the state classification network.
In a preferred embodiment of the present invention, since the neural network is only a relatively efficient method for expressing the non-linear mapping relationship, a lookup table having the same function as the neural network, a non-linear mapping function, or the like, can also be used to implement the required non-linear mapping relationship.
In the invention, the neural network control method takes the refrigerating capacity, the heating capacity or the energy consumption output by the traditional fuzzy control method or the PID control method as a target. Firstly, as shown in fig. 4, under the premise of ensuring the finally output cooling capacity or heating capacity, the energy consumption obtained by the neural network control method in the controllable range of the control capacity is not higher than that obtained by the conventional fuzzy control method or PID control method. Secondly, on the premise of ensuring the same or nearly the same energy consumption, in the controllable range of the control quantity, the refrigerating quantity or the heating quantity obtained by adopting the neural network control method is not lower than that obtained by adopting the traditional fuzzy control method or the PID control method. Thirdly, under the condition of a given measurement M, for example, M = a × cooling capacity (or heating capacity) -b × energy consumption, wherein a and b are coefficients, the measurement M obtained by the neural network control method is not lower than the measurement M obtained by the traditional fuzzy control method or the PID control method within the controllable control quantity range. That is, the neural network control method according to the present invention can have a better control effect, such as a higher cooling capacity or heating capacity, or less energy consumption, in the case of ensuring the same control target, as compared with the conventional control method.
The advantages of the neural network control method compared with the conventional control method will be described in detail below by using a specific embodiment. As shown in fig. 7, the state quantity is obtained by the conventional control method, and then the final control quantity is obtained by the neural network control method. However, the invention can update the control quantity by using the neural network control method without changing the traditional control method (fuzzy control method or PID control method) so as to achieve the purpose of saving energy consumption on the premise of generating the same refrigerating capacity. In fig. 7, the control amounts output by the conventional control method are: the rotating speed Fi =800 of the fan of the indoor unit, the rotating speed Fo =800 of the fan of the outdoor unit and the frequency F =70 of the compressor, the energy consumption generated correspondingly by the traditional control method is E =0.9, and the refrigerating capacity C = 3.4. Further, the neural network control method of the present invention is adopted to continuously update the control quantities, which outputs new control quantities according to the input control quantities Fi, Fo and F generated by the conventional control method as follows: the rotating speed Fi ' =900 of the fan of the indoor unit, the rotating speed Fo ' =900 of the fan of the outdoor unit and the frequency F ' =60 of the compressor, and after the updating through the neural network, the finally correspondingly generated energy consumption is E =0.7, which is reduced by 22% compared with the traditional control method, but the generated cooling capacity is still C =3.4 and has no influence. In this process, the observed quantity can be used as an optional input, and the control quantity obtained by the conventional control method and the current value and the historical value of the observed quantity can be used as the input of the neural network.
In summary, compared with the conventional control method, the neural network-based air conditioner control method provided by the invention has a better control effect under the condition of ensuring that the same control target is finally achieved, so that the air conditioner system has higher refrigerating capacity or heating capacity and less energy consumption.
While the present invention has been described in detail with reference to the preferred embodiments, it should be understood that the above description should not be taken as limiting the invention. Various modifications and alterations to this invention will become apparent to those skilled in the art upon reading the foregoing description. Accordingly, the scope of the invention should be determined from the following claims.
Claims (10)
1. An air conditioner control method based on a neural network is characterized by comprising the following steps:
the neural network control method is combined with the traditional control method, the observed quantity of the air conditioning system is automatically detected, and the generated control quantity acts on the air conditioning system to automatically control, so that the refrigerating capacity, the heating capacity or the energy consumption of the air conditioning system are always kept at expected values;
the neural network consists of a plurality of groups of neurons connected in parallel or in series, each neuron corresponds to an independent nonlinear mapping function, and the neural network can approximate any linear or nonlinear function; the input of the neural network comprises state quantity output by a traditional control method or observed quantity and/or feedback control quantity of the air conditioning system; wherein the observed quantity comprises the indoor and outdoor ambient temperature, the indoor and outdoor ambient humidity, and the surface temperature of the condenser and the evaporator; the output of the neural network is the controlled quantity or the new observed quantity after being controlled and adjusted by the neural network; wherein the control quantity comprises at least one or any combination of the following variables: compressor frequency, indoor unit fan speed, outdoor unit fan speed, and expansion valve opening;
according to the statistical data of the observed quantity of the air conditioning system, dynamically configuring and updating mapping functions of a plurality of neurons or mapping functions of the whole neural network;
wherein, the traditional control method comprises the following steps: a fuzzy control method or a PID control method;
the method for keeping the refrigerating capacity, the heating capacity or the energy consumption of the air conditioning system at a desired value specifically comprises the following steps: when the absolute value of the difference between the refrigerating capacity or the heating capacity of the air conditioning system and the refrigerating capacity or the heating capacity obtained by the traditional control method is smaller than a certain threshold value, the energy consumption of the air conditioning system is kept to be not higher than that of the traditional control method; and when the absolute value of the difference between the energy consumption of the air conditioning system and the energy consumption required by the traditional control method is smaller than a certain threshold, keeping the refrigerating capacity or the heating capacity of the air conditioning system not lower than that of the traditional control method.
2. The neural network-based air conditioner control method of claim 1, comprising the steps of:
SA1, adopting traditional control method, fuzzy control method or PID control method, to automatically detect the observed quantity of air-conditioning system, and outputting corresponding state quantity;
SA2, controlling and regulating the state quantity by adopting a neural network control method, and outputting a corresponding control quantity;
SA3, and automatically controlling the control amount by applying the control amount to the air conditioning system.
3. The neural network-based air conditioner control method according to claim 2, wherein the state quantity comprises at least one or any combination of the following variables: the control amount generated by the conventional control method; estimating the required refrigerating capacity or other parameters corresponding to the required refrigerating capacity based on the observed quantity; estimating other parameters based on the observed quantity; measuring the observed quantity; the amount of control fed back.
4. The neural network-based air conditioner controlling method of claim 2, wherein the state quantity includes a current state quantity and/or a historical state quantity.
5. The neural network-based air conditioner control method of claim 1, comprising the steps of:
SB1, adopting a neural network control method to control and regulate the observed quantity and/or the feedback control quantity of the air conditioning system, and outputting a new observed quantity;
SB2, adopting traditional control method, namely fuzzy control method or PID control method, to automatically detect the new observed quantity and output corresponding control quantity;
SB3, the control amount is applied to the air conditioning system to automatically control the air conditioning system.
6. The neural network-based air conditioner control method of claim 1, comprising the steps of:
SC1, fitting the traditional control method, namely fuzzy control method or PID control method, into the neural network control method with the same or similar refrigerating and/or heating capability and energy consumption;
SC2, adopting the neural network control method fitting the traditional control method, automatically detecting and controlling and adjusting the observed quantity and/or feedback control quantity of the air-conditioning system, and outputting the corresponding control quantity;
and SC3, applying the control quantity to the air conditioning system to automatically control.
7. The neural network-based air conditioner control method according to claim 6, wherein the air conditioner control method having similar cooling and/or heating capabilities and energy consumption specifically means:
when the absolute value of the difference between the refrigerating capacity or the heating capacity of the air conditioning system obtained by adopting the neural network control method fitted by the traditional control method and the refrigerating capacity or the heating capacity obtained by adopting the traditional control method is smaller than a certain threshold value, the energy consumption of the air conditioning system is kept to be not higher than that of the traditional control method;
and when the absolute value of the difference between the energy consumption of the air conditioning system obtained by adopting the neural network control method fitted by the traditional control method and the energy consumption required by the traditional control method is smaller than a certain threshold value, keeping the refrigerating capacity or the heating capacity of the air conditioning system not lower than that of the traditional control method.
8. The neural network-based air conditioner controlling method according to claim 2, 5 or 6, wherein the control amount is further differentially controlled.
9. The neural network-based air conditioner control method according to claim 2, 5 or 6, wherein the neural network adopts a cascade structure and is formed by cascading at least one basic network and at least one control quantity generating network;
the basic network carries out feature extraction on the input state quantity, the observed quantity and the feedback control quantity of the air conditioning system to obtain the intermediate quantity of the neural network, and the control quantity of the air conditioning system is output through the network generated by the control quantity.
10. The neural network-based air conditioner control method according to claim 2, 5 or 6, wherein the neural network adopts a hybrid structure and is formed by cascading a state classification network and a plurality of control quantity generation networks;
the state classification network classifies the input state quantity, observation quantity and feedback control quantity of the air conditioning system; and each control quantity generation network correspondingly outputs the control quantity of the air conditioning system according to the classification result of each type.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811155298.8A CN108895618B (en) | 2018-09-30 | 2018-09-30 | Air conditioner control method based on neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811155298.8A CN108895618B (en) | 2018-09-30 | 2018-09-30 | Air conditioner control method based on neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108895618A CN108895618A (en) | 2018-11-27 |
CN108895618B true CN108895618B (en) | 2021-11-26 |
Family
ID=64360206
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811155298.8A Active CN108895618B (en) | 2018-09-30 | 2018-09-30 | Air conditioner control method based on neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108895618B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109631238A (en) * | 2019-01-28 | 2019-04-16 | 宁波溪棠信息科技有限公司 | A kind of control system and control method improving air-conditioning system operational energy efficiency |
CN110094837B (en) * | 2019-04-30 | 2020-09-04 | 珠海格力电器股份有限公司 | Intelligent control device and method for air conditioner |
CN110736228B (en) * | 2019-10-29 | 2020-11-27 | 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 |
CN110736244B (en) * | 2019-10-29 | 2020-10-20 | 珠海格力电器股份有限公司 | Method and device for controlling air conditioner and air conditioner |
CN111520883B (en) * | 2020-04-29 | 2022-03-04 | 宁波溪棠信息科技有限公司 | Air conditioner control method and system based on target room load prediction |
CN111928420B (en) * | 2020-07-17 | 2022-02-11 | 珠海格力电器股份有限公司 | Air conditioner control method and device, electronic equipment and storage medium |
CN112346333B (en) * | 2020-11-05 | 2023-03-24 | 中国航空工业集团公司西安航空计算技术研究所 | Compressor rotating speed control method based on BP neural network regulation |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101498534A (en) * | 2008-12-08 | 2009-08-05 | 天津大学 | Multi-target intelligent control method for electronic expansion valve of refrigeration air conditioner heat pump system |
JP2010030499A (en) * | 2008-07-30 | 2010-02-12 | Denso Corp | Vehicular air-conditioner |
CN104089362A (en) * | 2014-06-03 | 2014-10-08 | 杭州哲达科技股份有限公司 | Cooling efficiency maximization method for cooling water system in central air-conditioner and control device |
-
2018
- 2018-09-30 CN CN201811155298.8A patent/CN108895618B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010030499A (en) * | 2008-07-30 | 2010-02-12 | Denso Corp | Vehicular air-conditioner |
CN101498534A (en) * | 2008-12-08 | 2009-08-05 | 天津大学 | Multi-target intelligent control method for electronic expansion valve of refrigeration air conditioner heat pump system |
CN104089362A (en) * | 2014-06-03 | 2014-10-08 | 杭州哲达科技股份有限公司 | Cooling efficiency maximization method for cooling water system in central air-conditioner and control device |
Also Published As
Publication number | Publication date |
---|---|
CN108895618A (en) | 2018-11-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108895618B (en) | Air conditioner control method based on neural network | |
Homod | Analysis and optimization of HVAC control systems based on energy and performance considerations for smart buildings | |
Homod et al. | Gradient auto-tuned Takagi–Sugeno Fuzzy Forward control of a HVAC system using predicted mean vote index | |
Pfeiffer et al. | Control of temperature and energy consumption in buildings-a review. | |
CN110057045A (en) | Control method for air conditioner | |
CN110059801A (en) | Energy efficiency of air conditioner control method neural network based | |
Li et al. | A novel neural network aided fuzzy logic controller for a variable speed (VS) direct expansion (DX) air conditioning (A/C) system | |
EP0405149A2 (en) | Room condition maintaining apparatus | |
CN110726218B (en) | Air conditioner, control method and device thereof, storage medium and processor | |
CN111520878A (en) | Air conditioner temperature control system based on RBF neural network and control method thereof | |
CN101498534A (en) | Multi-target intelligent control method for electronic expansion valve of refrigeration air conditioner heat pump system | |
Habib et al. | Fuzzy climate decision support systems for tomatoes in high tunnels | |
CN104833154B (en) | Chilled water loop control method based on fuzzy PID and neural internal model | |
Singh et al. | Fuzzy modeling and control of HVAC systems–A review | |
Wang et al. | Study of neural network PID control in variable-frequency air-conditioning system | |
CN112628956A (en) | Water chilling unit load prediction control method and system based on edge cloud cooperative framework | |
CN109140660A (en) | For the Intelligent temperature controller method and device of air-conditioning, air-conditioning, storage medium | |
CN114046593A (en) | Dynamic predictive machine learning type air conditioner energy-saving control method and system | |
CN111649457A (en) | Dynamic predictive machine learning type air conditioner energy-saving control method | |
CN111854063A (en) | Control method of variable frequency air conditioner | |
Becker et al. | Fuzzy control for temperature and humidity in refrigeration systems | |
Murzamadieva et al. | Development of a system for ensuring humidity in sport complexes | |
CN112346333B (en) | Compressor rotating speed control method based on BP neural network regulation | |
CN113821903A (en) | Temperature control method and device, modular data center and storage medium | |
CN116101496A (en) | Control method for energy efficiency optimization of electric environmental control system of airplane |
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 | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20220628 Address after: Room 201, block B, Changtai Plaza, Lane 2889, Jinke Road, Pudong New Area, Shanghai 200120 Patentee after: Hengxuan Technology (Shanghai) Co.,Ltd. Address before: 315500 room 701, No. 88, Dongfeng Road, Yuelin street, Fenghua District, Ningbo City, Zhejiang Province Patentee before: NINGBO XITANG INFORMATION TECHNOLOGY Co.,Ltd. |