Control method of air conditioner and related product

Abstract

The invention provides a control method of an air conditioner and a related product. The method comprises the following steps: acquiring operation data of the air conditioner; inputting the operation data into a deep reinforcement learning model to obtain factory setting parameters of the air conditioner; the deep reinforcement learning model is configured to change part or all factory setting parameters of the air conditioner; and running based on the factory setting parameters. Due to the fact that the air conditioner is provided with the deep reinforcement learning model, the deep reinforcement learning model is configured to change part of or all factory setting parameters of the air conditioner, the air conditioner can adapt to the complex installation environment and use environment, the application range of the air conditioner is widened, and the practicability of the air conditioner is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to air conditioning technology, and in particular, to a control method of an air conditioner and related products. BACKGROUND

[0002] With the development of society and the continuous improvement of people's living standards, various air conditioning devices have become one of the indispensable electrical appliances in people's daily life. Various air conditioning devices can help people reach an adaptable temperature when the environmental temperature is too high or too low. The current air conditioning devices mainly include various types of air conditioners. The existing air conditioners all store some control parameters, which are called factory setting parameters. The air conditioner will operate according to the factory setting parameters. The original factory setting parameters can be fixed parameters or adjustable parameters based on fixed parameters and compensation values. However, in complex installation and use environments, the original factory setting parameters of the air conditioner may not meet the current use scenarios of the user, such as outdoor unit exhaust air short circuit, resulting in site temperature rise, heat cannot be effectively exchanged in high temperature, etc. SUMMARY

[0003] In view of the above problems, the present application is proposed in order to provide a control method of an air conditioner and related products which overcomes the above problems or at least partially solves the above problems. The factory setting parameters of the air conditioner are adjusted based on the current running data of the air conditioner to meet the current installation and use environments.

[0004] According to one aspect of the present application, the present application provides a control method of an air conditioner, comprising:

[0005] obtaining running data of the air conditioner;

[0006] inputting the running data into a deep reinforcement learning model to obtain factory setting parameters of the air conditioner; the deep reinforcement learning model is configured to change part or all of the factory setting parameters of the air conditioner;

[0007] running based on the factory setting parameters.

[0008] Optionally, the part or all of the factory setting parameters of the air conditioner include the rotating speed of the outdoor fan, the opening degree of the electronic expansion valve, and the proportional coefficient, integral coefficient and differential coefficient of the PID control strategy; the PID control strategy is used to control the rotating speed of the compressor.

[0009] Optionally, the PID control strategy takes the difference between the indoor target temperature and the indoor environment temperature as input and takes the compressor rotating speed as output.

[0010] Optionally, the deep reinforcement learning model includes a state space, an action space and a reward function.

[0011] The state space comprises an outdoor heat exchanger heat exchange capacity, an outdoor temperature, an outdoor fan rotating speed, a compressor frequency, an evaporation temperature, a condensation temperature and an expansion valve opening degree;

[0012] The action space comprises an outdoor fan rotating speed adjustment step, a compressor frequency adjustment step and an electronic expansion valve opening degree adjustment step.

[0013] The reward function is constructed based on a refrigerating or heating capacity of the air conditioner, a power consumption of the compressor and a power consumption of the outdoor fan.

[0014] Optionally, the deep reinforcement learning model is a deep deterministic policy gradient model or a proximal policy optimization model.

[0015] Optionally, the operating data comprises an outdoor temperature, an outdoor fan rotating speed, a compressor frequency, an evaporation temperature, a condensation temperature and an expansion valve opening degree.

[0016] The outdoor heat exchanger heat exchange capacity is obtained based on the outdoor temperature, the outdoor fan rotating speed, the evaporation temperature and the condensation temperature.

[0017] Optionally, the outdoor temperature is an outdoor ambient temperature, or the outdoor temperature comprises an outdoor unit inlet air temperature and an outdoor unit outlet air temperature.

[0018] According to another aspect of the present application, there is also provided a computer readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the steps of any of the above air conditioner control methods.

[0019] According to still another aspect of the present application, there is also provided a computer program product comprising a computer program, the computer program being executed by a processor to implement the steps of any of the above air conditioner control methods.

[0020] According to yet another aspect of the present application, there is also provided an air conditioner comprising a memory, a processor and a computer program stored in the memory, the processor executing the computer program to implement the steps of any of the above air conditioner control methods.

[0021] The air conditioner control method and related products of the present application have a deep reinforcement learning model configured to change part or all of the factory setting parameters of the air conditioner, so that the air conditioner can adapt to complex installation and use environments, and the application range and practicality of the air conditioner are improved.

[0022] Further, the control method of the air conditioner and the related product can change and adjust the proportional coefficient, the integral coefficient and the differential coefficient of the PID control strategy, the rotating speed of the outdoor fan, the opening degree of the electronic expansion valve and the like, that is, the deep reinforcement learning is combined with the PID control, the outdoor fan rotating speed, the compressor frequency and the like can be adaptively adjusted, and the heat exchange efficiency and the refrigerating capacity can be significantly improved.

[0023] The above and other objects, advantages and features of the present application will become more apparent from the following detailed description of specific embodiments thereof, taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0024] Some specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. The same reference numbers in different drawings denote the same or similar components or parts. It should be understood by those skilled in the art that the drawings are not necessarily drawn to scale. In the drawings:

[0025] Figure 1 is a schematic flowchart of a control method of an air conditioner according to an embodiment of the present application;

[0026] Figure 2 is a schematic diagram of a computer program product according to an embodiment of the present application;

[0027] Figure 3 is a schematic diagram of a computer-readable storage medium according to an embodiment of the present application. DETAILED DESCRIPTION

[0028] The embodiment of the present application provides a control method of an air conditioner. Figure 1 is a schematic flowchart of a control method of an air conditioner according to an embodiment of the present application, which can generally include:

[0029] In step S101, operation data of the air conditioner is acquired.

[0030] In step S102, the operation data is input into a deep reinforcement learning model to obtain factory setting parameters of the air conditioner; the deep reinforcement learning model is configured to change part or all of the factory setting parameters of the air conditioner.

[0031] In step S103, the operation is performed based on the factory setting parameters.

[0032] In the control method of the air conditioner according to the embodiment of the present application, the deep reinforcement learning model can be built in the air conditioner. When the air conditioner is running, the operation data corresponding to the input of the deep reinforcement learning model is acquired, and based on the operation data, the deep reinforcement learning model outputs the factory setting parameters of the air conditioner. Then, the air conditioner is run based on the factory setting parameters output by the deep reinforcement learning model.

[0033] The control method of the air conditioner of the embodiment of the present application has a deep reinforcement learning model configured to change part or all of the factory setting parameters of the air conditioner, so that the air conditioner can adapt to complex installation and use environments, and the application range and practicality of the air conditioner are improved. Moreover, the air conditioner control method based on the deep reinforcement learning model can quickly respond in a complex environment, and the flexibility is improved, a more efficient and adaptive air conditioner control is realized, and the energy efficiency, indoor comfort and applicability of the system of the air conditioner are significantly improved.

[0034] In some embodiments of the present application, the part or all of the factory setting parameters of the air conditioner include the speed of the outdoor fan, the opening degree of the electronic expansion valve, and the proportional coefficient, integral coefficient and differential coefficient of the PID control strategy; the PID control strategy is used to control the speed of the compressor. The air conditioner can change and adjust the proportional coefficient, integral coefficient and differential coefficient of the PID control strategy, the speed of the outdoor fan, the opening degree of the electronic expansion valve, etc., that is, the deep reinforcement learning is combined with the PID control, the speed of the outdoor fan and the frequency of the compressor can be adaptively adjusted, the heat exchange efficiency and refrigeration capacity can be significantly improved, and the energy consumption can be reduced, and different environmental conditions can be adapted.

[0035] In some embodiments of the present application, the PID control strategy takes the difference between the indoor target temperature and the indoor environment temperature as the input, and takes the speed of the compressor as the output. The PID control strategy using the indoor target temperature and the indoor environment temperature can make the indoor temperature quickly and stably reach the target temperature and remain at the target temperature, forming a comfortable indoor environment and improving the comfort of the human body.

[0036] In some embodiments of the present application, the part or all of the factory setting parameters of the air conditioner also include the limit frequency of the compressor. In high-temperature weather, the limit frequency of the compressor can cause insufficient refrigeration capacity. The deep reinforcement learning model can change the limit frequency of the compressor, for example, to make the limit frequency of the compressor higher.

[0037] In some embodiments of the present application, the deep reinforcement learning model includes a state space, an action space and a reward function.

[0038] The state space includes the heat exchange capacity of the outdoor heat exchanger, the outdoor temperature, the speed of the outdoor fan, the frequency of the compressor, the evaporation temperature, the condensation temperature and the opening degree of the expansion valve. The deep reinforcement learning model fully considers the heat exchange capacity of the outdoor heat exchanger, which can adapt to the complex installation and use environments of the outdoor unit, such as outdoor high-temperature environment, outdoor unit exhaust short circuit, low-load refrigeration and other use or installation environments, and can significantly improve the coefficient of performance of the air conditioner.

[0039] Specifically, the conventional PID control usually uses fixed parameters, but in a complex outdoor unit installation environment, the fixed parameters can lead to poor control effect. For example: in a high-temperature environment, the heat exchange capacity decreases, causing the compressor frequency to rise, but the outdoor fan speed fails to effectively match, and the heat exchange efficiency decreases. In the case of outdoor fan short-circuit, high-speed outdoor fan can cause hot air backflow, affecting heat dissipation. In the embodiment of the present application, when the heat exchange capacity decreases, the outdoor fan speed can be increased through the deep reinforcement learning model to enhance the condenser heat dissipation effect. In the low heat exchange capacity scenario, the differential coefficient of the compressor frequency control can be increased through the deep reinforcement learning model to avoid frequent fluctuations caused by rapid temperature rise.

[0040] The action space includes the adjustment step of the outdoor fan speed, the adjustment step of the compressor frequency, and the adjustment step of the opening degree of the electronic expansion valve.

[0041] The deep reinforcement learning model constructs a reward function based on the refrigerating capacity or heating capacity of the air conditioner, the power consumption of the compressor, and the power consumption of the outdoor fan. The reward function constructed in the embodiment of the present application fully considers the relationship between the refrigerating capacity or heating capacity and the power consumption of the compressor and the power consumption of the outdoor fan, so that the higher the refrigerating capacity is, the better, and the lower the power consumption of the compressor and the outdoor fan is, the better.

[0042] Specifically, the reward function can be set as: R=Q cooling -α(P comp +P fan ). Wherein, p comp is the power consumption of the compressor, P fan is the power consumption of the fan, α is the power consumption weight coefficient, Q cooling is the refrigerating capacity or heating capacity. The refrigerating capacity or heating capacity can be calculated based on the evaporation temperature and the condensation temperature.

[0043] In some embodiments of the present application, the deep reinforcement learning model is a deep deterministic policy gradient model. The deep deterministic policy gradient (DDPG) algorithm is a deep reinforcement learning algorithm suitable for solving continuous action space problems. It combines deterministic policy and deep neural network, is a model-independent reinforcement learning algorithm, belongs to the Actor-Critic framework, and simultaneously utilizes the advantages of DQN and PG (Policy Gradient). In the embodiment of the present application, the deep deterministic policy gradient model can continuously adjust the opening degree of the electronic expansion valve, the speed of the compressor, and the speed of the outdoor fan.

[0044] In other embodiments of the present invention, the deep reinforcement learning model is a proximal policy optimization model. Proximal Policy Optimization (PPO) is a reinforcement learning algorithm designed to address the problems of training instability and low sample efficiency in deep reinforcement learning. The PPO algorithm is based on policy gradients and trains an agent by optimizing the policy to maximize long-term rewards. Compared to other algorithms, PPO has the advantages of simplicity, efficiency, and stability. PPO improves the training process through two key concepts: proximal policy optimization and a clipping objective function. Proximal policy optimization maintains training stability by limiting the size of policy updates, ensuring that each update is within an acceptable range. The clipping objective function is the core concept of the PPO algorithm. When updating the policy, it uses the clipping objective function to constrain the amplitude of the policy update to avoid excessive updates that lead to training instability. The PPO algorithm uses the proximal policy optimization method to optimize the policy to avoid the problem of performance degradation caused by overly aggressive policy updates. It can ensure the performance of the air conditioner and continuously adjust the opening of the electronic expansion valve, the speed of the compressor, and the speed of the outdoor fan.

[0045] In some embodiments of the present invention, the operating data includes outdoor temperature, outdoor fan speed, compressor frequency, evaporation temperature, condensation temperature and expansion valve opening.

[0046] The heat exchange capacity of the outdoor heat exchanger is obtained based on the outdoor temperature, the speed of the outdoor fan, the evaporation temperature and the condensation temperature. The heat exchange capacity of the outdoor heat exchanger is usually a heat exchange capacity coefficient, which can be obtained by the ratio between the actual heat exchange rate and the theoretical heat exchange rate. The theoretical heat exchange rate can be obtained by looking up the relationship table between the current outdoor temperature, the speed of the outdoor fan and the theoretical heat exchange rate. The actual heat exchange rate can be calculated based on the current evaporation temperature and the condensation temperature. Furthermore, the outdoor temperature is the outdoor ambient temperature. Alternatively, in other embodiments of the present invention, the outdoor temperature includes the outdoor unit inlet air temperature and the outdoor unit outlet air temperature. The heat exchange capacity of the outdoor heat exchanger can be used to determine whether the current heat exchange capacity is limited, thereby triggering the control strategy of the deep reinforcement learning model.

[0047] An air conditioner using the control method for an air conditioner according to an embodiment of the present invention has superior technical effects compared to existing air conditioner control methods. Existing air conditioners use traditional PID control, where the proportional coefficient, integral coefficient, and differential coefficient of the PID control strategy are fixed parameters, and the outdoor fan speed is also a fixed parameter.

[0048] For example, under high temperature conditions, the test environment is an outdoor temperature of 40°C and an air humidity of 60% RH. The experimental data comparison is shown in Table 1:

[0049] Table (1)

[0050]

[0051] Under the traditional PID control, the fan speed is fixed, and the heat exchange capacity is reduced, resulting in low refrigeration capacity and EER. The control method of the air conditioner of the embodiment of the application increases the compressor frequency by improving the outer fan speed, optimizing the opening degree of the electronic expansion valve, and changing the proportional coefficient, integral coefficient and differential coefficient of the PID control strategy through the deep reinforcement learning model. Obviously, the heat exchange capacity is increased by 13.3%, the refrigeration capacity is increased by 10.8%, and the EER is increased by 14.3%.

[0052] For example, under the exhaust air short circuit condition, the test environment is that the outdoor unit is closed around and only a 30cm air outlet channel. The experimental data comparison is shown in Table (II):

[0053] Table (II)

[0054]

[0055] Under the traditional PID control, the problem cannot be identified due to the exhaust air short circuit, the outer fan runs at high speed but the effect is not good, the condensing temperature rises, and the heat exchange capacity is reduced. The control method of the air conditioner of the embodiment of the application reduces the compressor frequency by reducing the outer fan speed, optimizing the opening degree of the electronic expansion valve, and changing the proportional coefficient, integral coefficient and differential coefficient of the PID control strategy through the deep reinforcement learning model. Obviously, the influence of the hot air backflow is reduced, the heat exchange capacity is increased by 17.6%, and the EER is increased by 24%.

[0056] For example, under the low load condition, the test environment is that the outdoor temperature is 28℃ and the room temperature is 24℃. The experimental data comparison is shown in Table (III):

[0057] Table (III)

[0058]

[0059] Under the traditional PID control, the energy consumption cannot be reduced under the low load condition, resulting in redundant operation of the compressor and the fan. The control method of the air conditioner of the embodiment of the application reduces the compressor frequency by reducing the outer fan speed, optimizing the opening degree of the electronic expansion valve, and changing the proportional coefficient, integral coefficient and differential coefficient of the PID control strategy through the deep reinforcement learning model. Unnecessary energy consumption can be reduced, the EER is increased by 14.3%, and the refrigeration capacity is not significantly reduced, reducing unnecessary energy consumption.

[0060] In summary, through experimental verification, the control method of the air conditioner is superior to the traditional PID control under various complex working conditions. The control method of the air conditioner can make the air conditioner dynamically adapt to the installation environment of the outdoor unit (high temperature, exhaust air short circuit, etc.), improve the heat exchange capacity, deeply strengthen the learning self-adaptive optimization, surpass the traditional PID fixed control, improve the energy efficiency ratio, be suitable for different load working conditions, ensure the efficient operation of the air conditioning system, and be widely suitable for high-efficiency energy-saving air conditioners, and can be popularized to household / commercial / industrial air conditioning systems.

[0061] The flowchart provided by the embodiment is not intended to indicate that the operations of the method should be performed in any particular order, or that all of the operations of the method are included in each case. In addition, the method can include additional operations. Additional changes can be made to the above method within the scope of the technical idea provided by the embodiment method.

[0062] It should be understood that in some embodiments, parts can be realized by hardware, software, firmware or a combination thereof. In the above implementation, a plurality of steps or methods can be realized by software or firmware stored in a memory and executed by a suitable instruction execution system.

[0063] The embodiment also provides a computer program product 10 and a computer readable storage medium 20. Figure 2 is a schematic diagram of the computer program product 10 according to an embodiment of the application, Figure 3 is a schematic diagram of the computer readable storage medium 20 according to an embodiment of the application. The computer program product 10 includes a computer program 11, which, when executed by the processor 32, realizes the steps of the control method of the air conditioner according to any of the above embodiments. The computer program 11 is stored on the computer readable storage medium 20, and when executed by the processor 32, realizes the steps of the control method of the air conditioner according to any of the above embodiments.

[0064] The computer program 11 for performing the operations of the present application can be in assemblies instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, configuration data for integrated circuits, or in source code or object code written in any combination of one or more programming languages, and process- programming languages. The computer program 11 can execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer can be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection can be made to an external computer, for example, through the Internet. In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate array (FPGA), or programmable logic array (PLA) can execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present application.

[0065] For the description of the present embodiment, the computer program product 10 is a product of manufacture that includes the computer program 11.

[0066] For the description of the present embodiment, the computer readable storage medium 20 is a tangible device that can retain and store computer program 11, which can be any medium (non-transitory) that can contain, store, communicate, propagate or transport the program 11 for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer readable storage medium 20 include the following: portable computer diskette, hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or Flash memory), static random access memory (SRAM), portable compact disc read-only memory (CD-ROM), digital versatile disk (DVD), memory stick, a floppy disk, a mechanical encoding device, and any suitable combination of the foregoing. A computer readable storage medium 20, a computer readable storage device, or a computer readable storage can be any available medium or device that can be accessed by a computer.

[0067] The present embodiments also provide an air conditioner. The air conditioner can include a memory, a processor, and the above computer program 11 stored on the memory and running on the processor. The processor is adapted to execute the stored instructions, and the processor can be a single core processor, a multi core processor, a computing cluster, or any number of other configurations. The memory provides temporary storage of the operations of the instructions during operation. The memory can include random access memory (RAM), read only memory, flash memory, or any other suitable memory system. The computer program 11, when executed by the processor, implements the steps of the control method of the air conditioner of any of the above embodiments.

[0068] At this point, those skilled in the art will fully appreciate that the application need not be limited to the exact details of construction, arrangement of components, and operation of the illustrative embodiments as set forth herein, and that many other variations and modifications can be made without departing from the spirit or scope of the application. Accordingly, the scope of the application should be judged in terms of the claims that follow.

Claims

1. A control method of an air conditioner, characterized by, The method comprises: obtaining operation data of the air conditioner; inputting the operation data into a deep reinforcement learning model to obtain factory setting parameters of the air conditioner; the deep reinforcement learning model is configured to change part or all of the factory setting parameters of the air conditioner; running based on the factory setting parameters.

2. The control method of the air conditioner according to claim 1, wherein the part or all of the factory setting parameters of the air conditioner include a rotating speed of an outdoor fan, an opening degree of an electronic expansion valve, and a proportional coefficient, an integral coefficient and a differential coefficient of a PID control strategy used to control a rotating speed of a compressor.

3. The control method of the air conditioner according to claim 2, wherein the PID control strategy takes a difference between an indoor target temperature and an indoor environment temperature as an input and takes the rotating speed of the compressor as an output.

4. The control method of the air conditioner according to claim 2, wherein the deep reinforcement learning model includes a state space, an action space and a reward function; the state space includes an outdoor heat exchanger heat exchange capacity, an outdoor temperature, an outdoor fan rotating speed, a compressor frequency, an evaporation temperature, a condensation temperature and an expansion valve opening degree; the action space includes an adjustment step of the outdoor fan rotating speed, an adjustment step of the compressor frequency and an opening degree adjustment step of the electronic expansion valve; the reward function is constructed based on a refrigeration or heating capacity of the air conditioner, a power consumption of the compressor and a power consumption of the outdoor fan.

5. The control method of the air conditioner according to claim 4, wherein the deep reinforcement learning model is a deep deterministic policy gradient model or a proximal policy optimization model.

6. The control method of the air conditioner according to claim 4, wherein the operation data includes an outdoor temperature, an outdoor fan rotating speed, a compressor frequency, an evaporation temperature, a condensation temperature and an expansion valve opening degree; the outdoor heat exchanger heat exchange capacity is obtained based on the outdoor temperature, the outdoor fan rotating speed, the evaporation temperature and the condensation temperature.

7. The control method of the air conditioner according to claim 6, wherein the outdoor temperature is an outdoor environment temperature, or the outdoor temperature includes an outdoor unit inlet air temperature and an outdoor unit outlet air temperature. The computer program is executed by a processor to implement the steps of the control method of the air conditioner according to any one of claims 1 to 7. The computer program is executed by a processor to implement the steps of the control method of the air conditioner according to any one of claims 1 to 7. The computer program is executed by a processor to implement the steps of the control method of the air conditioner according to any one of claims 1 to 7. ​ ​ ​ 8. A computer-readable storage medium having stored thereon a computer program, characterized in that, ​ 9. A computer program product comprising a computer program, characterized in that, ​ 10. An air conditioner characterized by comprising: ​

Images