KR0134727B1 - Method for moving an airconditioner - Google Patents

Abstract

The present invention relates to an artificial intelligence driving method of an air conditioner, and more particularly, to an artificial intelligence driving method of an air conditioner for more convenience by reducing the number of operations by learning a user's air conditioning usage habits.

The conventional AI operation of the air conditioner has a problem in that it does not flexibly respond to the user's characteristics (ie, constitution or use characteristics) in order to control the operation with PMV = 0 in the case of PMV operation. Hot or cold key was adopted to solve the problem, but there was also a problem that the user is inconvenient due to repeated key operation for repeated environment (meal or cleaning).

Therefore, the present invention determines the control variable value corresponding to the new comfort index by moving the comfort index value for the selection key and the situation in which the key is pressed every time the selection key (hot / cold key) is inputted, Operon generation process to generate and store operon and search for genes in current time during operation to determine the control variable value that the comfort indicator becomes zero when there is no gene. During the driving control process and the gene present in the present time gene search result, the Euclidean distance and the expression intensity between the current environment and the transcription factor are calculated to determine the value of the control variable and when the gene exists to control the operation according to the new control variable value. The operation control process is executed sequentially to solve the above problems of the prior art.

Description

AI operation method of air conditioner

1 is a conceptual diagram of a conventional artificial intelligence air conditioner.

2 is a conceptual diagram of an artificial intelligence air conditioner to which the present invention is applied.

3 is an explanatory diagram for explaining the generation algorithm applied to the present invention.

4 is an explanatory diagram for explaining a control process according to the present invention.

Figure 5 is a diagram explaining the learning content by the development algorithm applied to the present invention.

6 is a signal flow diagram of the artificial intelligence driving process of the present invention air conditioner.

The present invention relates to an artificial intelligence driving method of an air conditioner, and more particularly, to an artificial intelligence driving method of an air conditioner for promoting convenience by reducing the number of operations by learning a user's air conditioning usage habits.

In the conventional air conditioner, the artificial intelligence operation calculates a comfort index in the neural network based on data such as suction temperature, outside air temperature, set temperature, air volume, and wind direction, as shown in FIG.

In this way, when the comfort index is calculated, the artificial intelligence operation is performed by controlling the air conditioner by changing the set temperature, air volume, and wind direction so that the comfort index (hereinafter, abbreviated as PMV) becomes 0.

In the above, PMV is a comfort indicator. When the room has a certain temperature, humidity, air flow, radiation environment, personal wear and activity, the evaluation of this complex environment is calculated by calculating PMV = -3 (cold), PMV =- To index of 2 (slightly cold), PMV = -1 (cool), PMV = 0 (normal), PMV = + 1 (warm), PMV = + 2 (slightly hot), PMV = + 3 (hot) This shows the comfort status of the current occupants.

In the case of such PMV control, it controls so that PMV = 0 in any complex environment.

In the case of PMV, PMV = 0 was set as a common response to a constant environment as an experiment on a constant human sample.

On the other hand, it is hot when you want to express the desire to make a difference in the average even in the same environment due to personal dissatisfaction with PMV = 0 and lifestyle differences (ie, patterns such as eating or cleaning) during the PMV control driving as described above. And there is a cold key, if you press the hot key once during the PMV control to operate the current temperature drops, press the cold key once to increase the current temperature to operate.

However, the AI operation of the conventional air conditioner has a problem in that the PMV = 0 does not flexibly respond to the user's characteristics (ie, constitution or use characteristics) to control the operation with PMV = 0.

In addition, the hot or cold key is employed to solve the complaint about the artificial intelligence driving, but there is a problem that the user is inconvenient due to the repeated key operation for the repeated environment (meal or cleaning).

Therefore, the present invention is to solve the above problems of the conventional air conditioner, the object of the present invention is to learn the air conditioning usage habits of the user to reduce the number of operations by artificial intelligence operating method of the air conditioner for more convenience In providing.

In order to achieve the object of the present invention, whenever a selection key (hot / cold key) is inputted, the control parameter value corresponding to the new comfort index is moved by moving the comfort index value for the selection key and the situation in which the key is pressed. Operon generation process to generate and store operon based on the new variable, and to search for the current time gene during operation and to determine the control variable value that the comfort index becomes zero when there is no gene. The operation control process for the absence of genes and the current time gene search results determine the control variable value by calculating the Euclidean distance and the expression intensity between the current environment and the transcription factor in the presence of the gene and determine the control variable value according to the new control variable value. If there is a gene that controls the operation, it consists of the operation control process.

Hereinafter, described in detail with reference to the accompanying drawings of the present invention.

2 is a system conceptual diagram of an artificial intelligence air conditioner according to the present invention. As shown in the drawing, it is comfortable by learning in a BP (Back Propagation) neural network based on data such as suction temperature, set temperature, air volume, wind direction, and suction temperature gradient. The index (PMV) is calculated and the new set temperature, air volume, and wind direction are determined according to the occurrence of the selection key (hot / cold key) in the generation algorithm with reference to the obtained comfort index (PMV).

Figure 3 is an explanatory diagram for explaining the developmental algorithm (HA) applied to the present invention, the gene is composed of DNA, the protein generation for the survival and inheritance of the body is the genetic information contained in the transfer region (B) By copying through RNA.

However, such copying is not always performed, but there is a function of controlling only the necessary information of a large number of genetic information to be properly expressed when necessary, which is collectively referred to as the promoter region (A).

The generation algorithm models the hierarchical structure of the actual control based on this and controls the variables to be controlled in a timely manner.

Hereinafter, the operations and effects of the present invention will be described with reference to FIGS. 4 to 6.

First, when the artificial intelligence function of the air conditioner is selected, the initial comfort index (PMV) is calculated to drive PMV = 0, and the search for the expression of the selection key (hot / cold) (S1).

In this case, if the selection key (hot / cold) is expressed, the time added by comparing the current time ti and the time to sum the preset time Δt to the current time ti is increased (S2). The next step is to search for the gene of the current time (S3).

As shown in (a) of FIG. 4, the gene is composed of elapsed time, current temperature, set temperature, air volume, wind direction, and suction temperature slope, and the transfer region is composed of set temperature, air volume, and wind direction.

Therefore, when the gene search result gene is present (S4) to calculate the adaptation time to the next step (S5).

Thereafter, when the calculated adaptation time is 0, a control variable value corresponding to the comfort index PMV is determined in the next step (S6-S9).

Gene (OPERON) is generated based on the determined comfort index and control variable value and stored in the memory of the microcomputer (S10).

That is, during the day, the user operates the select key (hot / cold key) several times according to the situation (meal, cleaning, etc.), and repeats the above steps (S2-S10) every time to control the environment and control at that time. Store the variable value in memory.

In the above memory, as shown in (b) of FIG. 4, a large number of gene groups (promoter region + transcription region) are formed (pop1-popn). Thus, the air conditioner operation of the first day is controlled as shown in (a) of FIG. Done.

Subsequently, when the air conditioner is operated the next day, as shown in FIG. 5B, the current gene is searched for at the elapsed time S11 (S12), and the presence or absence of the gene is determined (S13).

If the genetic search result is in agreement with yesterday's situation, the user's key operation is predicted, and the control index that the comfort index (PMV) becomes 0 is searched through the comfort index (PMV) table without the user's key manipulation. The value is determined, and the control is performed so that the comfort index PMV becomes 0 according to the determined control variable value, thereby performing the automatic operation of the artificial intelligence of the air conditioner (S14-S17).

The above control compares yesterday's time zone with today's time zone, detects the PMV of the corresponding time zone, sets the control variable value at each time, and sets the control index value according to the set control variable value. Operation is controlled to be zero.

On the other hand, if the gene search result (S12) gene is present to calculate the Euclidean distance (Ti) between the current environment and the transcription factor (S18).

The Euclidean distance (Ti) is the same as the Hamming Distance (Harming Distance).

Euclidean distance Ti is calculated as (S18).

Thereafter, when the dresshold or space radial (To) is increased by comparing the calculated Euclidean distance Ti and the dresshold or space radial, the expression intensity by the comfort indicator PMV is calculated (S20).

In this case, the expression intensity calculation is performed when the value of the comfort indicator (PMV) is 1 or less, that is, 0.5, for example, to control only a part of the set temperature, the air volume, and the wind direction.

In this way, when the calculation of the expression intensity is completed, the electronic start point is determined (S21), and the control variable value of the transfer area is copied to the present value (S22) to control the operation of the air conditioner.

On the other hand, if the air conditioner operates the next day when the air conditioner is different from yesterday's situation (for example, if the outside temperature suddenly changes drastically, winter → spring, spring → summer, etc.), the copy of the gene group is suppressed, and this time corresponds to the key operation. In this way, the predictive control is suppressed and the current control situation is maintained as shown in FIG.

As described in detail above, the present invention can learn the air conditioner use habits of the in-patients can control the air conditioner according to the use habits in advance prediction without the key operation of the occupants can eliminate the inconvenience caused by the cumbersome key operation of the existing user It works.

Claims (3)

In the artificial intelligence operation method of the air conditioner employing the artificial intelligence control method by the Generic Algorithm, each time a selection key (hot / cold key) is input, the situation of pressing the key and the comfort index value for the selection key To determine the value of the control variable corresponding to the new comfort index, and to generate and store the operon based on the new variable, and to search for the current time gene during operation, Calculate the driving control process in the absence of the gene to control the driving so that the comfort indicator becomes 0 by determining the control variable value, and calculate the Euclidean distance and the expression intensity between the current environment and the electronic factor in the presence of the gene. To determine the value of the control variable and to control the operation according to the new control variable value. Artificial intelligence operation method of an air conditioner that. The method of claim 1, wherein the generating of the operon comprises: searching for a current time gene and determining whether a gene exists when the selection key is expressed; and calculating the adaptation time when the gene is present and calculating the adaptation time. If it is not 0, returning to the selection key determining step, and if the adaptation time is 0, shifting the comfort index, comparing the shifted comfort index value with a predetermined comfort index table, and determining a control variable value; Artificial intelligence operating method of the air conditioner comprising the step of generating and storing a gene corresponding to the control variable value. The method of claim 1, wherein the driving control process in the presence of the gene comprises: searching for the presence of the gene during operation, calculating a Euclidean distance between the current environment and a transcription factor when the gene is present, and the calculated Euclidean distance; Comparing (Ti) with the dresshold or space radial value and returning to the first step when the Euclidean distance is large, and calculating the expression intensity by the comfort indicator when the Euclidean distance is small, and the calculated Determining a transcription start point according to the expression intensity, determining a control variable value corresponding to the calculated expression intensity, and controlling the air conditioner according to the determined control variable value. Intelligent driving method.