KR102540899B1 - Intelligent automatic feeding system for fish farms using artificial intelligence - Google Patents

Abstract

The present invention relates to an apparatus and method for supplying feed to fish in a farm. Artificial intelligence (AI) solves the problem of the supply method and the problem of the automatic feeders of the prior art, which had the problem of not being able to administer the optimal supply amount for fish growth due to simply supplying feed only according to the predetermined time and feeding amount. ) automatically determines the appropriate amount of feed depending on the fish species, growth or activity, and evenly supplies the required amount of feed throughout the farm. By being configured to remotely check and control the feeding process through a smartphone or computer, it is possible to solve all the disadvantages of the existing manual feeding method and automatic feeding machines, and at the same time, remote control is possible even if you are not on site. Provided is an intelligent automatic feeding system for fish farms using artificial intelligence configured to ensure accurate feeding through

Description

Intelligent automatic feeding system for fish farms using artificial intelligence}

The present invention relates to a feed supply device and method for supplying feed to fish raised in a farm, and more particularly, when supplying feed in a farm, a conventional manual feeding method in which a farm manager directly sprays feed A lot of labor is required, and because the farm manager arbitrarily pays based on his experience without considering the size of the farm or the difference according to the fish species and the feeding time, the proper amount of feed cannot be maintained and feed is wasted in many cases. , In order to solve the problem of the prior art farm feed supply method, which had a problem that the feed was not evenly distributed throughout the farm due to manual spraying, which hindered the uniform growth of farmed fish, artificial intelligence (AI) ), it is configured to accurately determine the appropriate amount of feed depending on the fish species, growth or activity of farmed fish, and evenly supply the required amount of feed throughout the farm. It relates to an intelligent automatic feeding system for fish farms using artificial intelligence that is configured to solve both problems and problems caused by the manual feeding method of feed.

In addition, the present invention, conventional automatic feeding machines for solving the problems of the manual feeding method as described above can save labor compared to manual feeding, but accurate accuracy based on the growth or activity of the fish species. Because it is configured to supply feed only according to the time and feeding amount pre-determined by the user without grounds, inaccurate feed supply is intensively administered to only a part of the entire farm, which hinders the even growth of all fish in the farm and leads to food competition. In order to solve the problems of conventional automatic feeders, which had a problem of causing water pollution due to oversupply of feed, in addition to injury to fish by artificial intelligence (AI), fish species and It is configured to accurately determine the appropriate feeding amount according to growth or activity and automatically supply the required amount of feed throughout the farm, thereby solving the disadvantages of both the existing manual feeding method and automatic feeding machines. It relates to an intelligent automatic feeding system for fish farms using artificial intelligence configured to

In addition, the present invention, as described above, is a conventional manual feeding method and automatic feed by more precisely and accurately supplying the amount of feed required according to the fish species, growth, activity, etc. of farmed fish through artificial intelligence (AI). At the same time, it is possible to solve all the disadvantages of feeders, and at the same time, various information on feed supply and overall aquaculture work obtained through artificial intelligence (AI) can be accessed remotely through mobile or web via a user's smartphone, tablet PC or computer. An intelligent automatic feeding system for fish farms using artificial intelligence configured to check and control the feeding process, so that accurate and appropriate feeding can always be performed through remote control even if the user is not at the farm site. will be.

In general, in order to farm fish in a farm, a certain amount of feed must be supplied for a certain amount of time.

In addition, in the past, in general, there were many cases in which feed was supplied by a manual supply method in which a person carries a feed container and sprays it on the farm, but this manual supply method requires a lot of labor, in addition to the size of the farm and the type of fish. In most cases, the farm manager arbitrarily sprayed feed based on his or her own experience, without considering the difference in feeding time and feeding time.

Therefore, in the conventional manual feeding method, because the manager arbitrarily spreads the feed depending only on his/her own experience, the optimum feeding amount for each fish species is not supplied and feed is wasted, or the feed is not evenly distributed throughout the farm, resulting in damage to the farmed fish. There was a problem that hindered uniform growth.

Accordingly, in recent years, in order to solve the problems of the existing manual feeding method as described above, an automatic feeding machine configured to automatically supply feed according to a predetermined amount has been widely used.

Here, as an example of the prior art for the feed supply device and method as described above, there is, for example, a "fish farm automatic feed feed device" as suggested in Korean Patent Registration No. 10-1620587.

More specifically, the aforementioned Korean Patent Registration Publication No. 10-1620587 discloses a main body having a discharge space formed on the front surface of the housing, a hopper in which feed is stored and a discharge port penetrated at the bottom, and feed on one side of the lower end of the hopper. A feed storage unit with a sensor for detecting the stored amount, a storage casing with input and output ports passing through the top and bottom to communicate with the feed storage unit, and installed to rotate by a pivoting shaft inside the storage casing to discharge the fixed amount of feed A blocking bucket with a concave inlet formed at an angle of 60 to 90 ° in the catalog storage space, a driving means fastened to one side of the rotation shaft to rotate the rotation shaft, and a discharge port provided to pass through the upper end of the discharge space of the body housing from the lower part of the outlet It is installed inside the discharge space of the main body and includes a scattering unit formed by a rotary spreader that rotates centrifugally by a driving means and a vertical rotation shaft to spread the feed discharged from the discharge port, setting a set amount of pellet feed for fish farming It relates to an automatic feed supply system for a fish farm configured to automatically spray on time.

In addition, another example of the prior art for the feed supply device and method as described above is, for example, a "feed supply device for a fish farm" as presented in Korean Patent Publication No. 10-2013-0000145.

More specifically, the above Korean Patent Publication No. 10-2013-0000145 discloses a hopper for accommodating feed, a rotating plate located at the bottom of the hopper and spraying feed falling from the hopper by a rotational force, rotating the rotating plate Including a power unit that allows the user to set the operation time, including a controller that applies power for the operation time to the power unit, reducing manpower consumption by automatically supplying the feed to the fish farm at regular intervals. It relates to a feed supply device for fish farms configured to prevent injuries due to competition for food.

As described above, various technical details have been proposed regarding automatic feeding devices for fish farms, but the contents of the prior art as described above have the following problems.

In other words, most of the automated fish feeding devices currently used in Korea supply a predetermined amount of feed to the farm at a specific time. Through this, it is possible to reduce the problem of labor shortage due to the decrease in the fisherman population due to aging and the expensive farm management labor cost.

On the other hand, most of the automatic feeding devices of the prior art as described above are simply configured to supply feed only according to the preset time and supply amount, thereby reducing the expensive feed amount, which is a decisive factor in aquaculture productivity. It is very difficult to properly adjust intelligently.

More specifically, the adequacy of feed supply is very important for the growth of aquaculture species and the improvement of the economic feasibility of farms, especially in tank-type land-based aquaculture, and aquaculture using existing live feed has emerged as a major cause of disease, and legal Due to the blocking of live feed, land-based fish farms supply artificially produced, expensive feed by manpower or by using an automatic feeding device.

Here, since the current workforce at the farm is very aging and the labor cost is very high, it is a very important key factor in aquaculture to properly supply expensive feed and increase feed efficiency, but the currently used automated fish feeding device is a method of supplying a certain amount of feed to the tank at a specific time and time, which can reduce expensive labor costs, but the amount of feed is not sufficient for the fish's growth environment (temperature, dissolved oxygen, etc.) and activity status (hunger, fullness, etc.) Regardless, there is a problem of under- or over-supply of feed because it is determined by the experience of the farm manager.

That is, if the feed is under-supplied, the growth of the fish does not work well, and if the feed is over-supplied, most of the feed remains in the tank as remnants and causes decay, resulting in water pollution and resulting disease in fish.

However, the automatic feeding devices of the prior art as described above are only focused on supplying feed in a predetermined amount at a predetermined time, and as described above, by determining the optimal supply amount in each farm environment A specific configuration for automatically supplying feed has not been suggested.

Therefore, in order to solve the problems of the automatic feeding devices of the prior art as described above, for example, by using artificial intelligence (AI), fish species of farmed fish and current growth environment information and real-time activity status The optimal supply amount is calculated based on the information such as fish farm, etc., and the process of supplying the necessary amount of feed evenly throughout the farm is automatically performed, maximizing the efficiency of feed supply while maintaining the growth rate of fish in the farm. A new configuration that can solve all the problems of the existing manual feeding method and automatic feeding devices, and at the same time, remote control is possible so that accurate and appropriate feeding can always be achieved through remote control even if the user is not at the farm site. It is desirable to present an automatic feed supplying device and method, but a device or method that satisfies all such demands until turnover has not been provided.

Korean Registered Patent Publication No. 10-1620587 (2016.05.04.) Korean Patent Publication No. 10-2013-0000145 (2013.01.02.)

The present invention is intended to solve the problems of the prior art as described above, and therefore, an object of the present invention is, without considering the size of the farm or the difference according to the fish species and the feeding time, etc., the farm manager arbitrarily depending on his experience In addition to the fact that due to feeding, the proper amount of feeding cannot be maintained and a lot of labor is required, the feed is not evenly distributed throughout the farm due to manual spraying, which hinders the uniform growth of farmed fish. In order to solve the problem of the method, artificial intelligence (AI) accurately determines the appropriate amount of feed depending on the fish species, growth, and activity of farmed fish, and automatically supplies the necessary amount of feed throughout the farm. It is intended to provide an intelligent automatic feeding system for fish farms using artificial intelligence, which is configured to solve both the problem of labor shortage due to the decrease in the fisherman population and the problem caused by the manual supply method of feed. .

In addition, another object of the present invention is that it is possible to reduce labor compared to the manual feeding method, but it is configured to supply feed only according to the time and amount of feeding simply predetermined by the user without accurate grounds based on the growth or activity of fish species. As a result, inaccurate feed supply is intensively administered to only a part of the entire farm, which hinders the even growth of all fish in the farm, and damages the fish due to food competition, as well as water pollution and disease due to oversupply of feed. In order to solve the problem of automatic feeding machines of the prior art, which had problems causing the occurrence of fish, artificial intelligence (AI) accurately determines the appropriate feeding amount according to the fish species, growth or activity of farmed fish, and provides the necessary amount of feed. Provides an intelligent automatic feeding system for fish farms using artificial intelligence, which is configured to automatically perform the processing of feeding evenly throughout the farm, thereby solving the disadvantages of both the existing manual feeding method and automatic feeding machines It's what you want to do.

In addition, another object of the present invention, as described above, is to more precisely and accurately supply the amount of feed required according to the fish species and growth or activity of farmed fish through artificial intelligence (AI), which is the conventional manual feed supply. It is possible to solve all the disadvantages of feeding methods and automatic feeding machines, and at the same time, various information on feeding and overall farming operations obtained through artificial intelligence (AI) can be transferred to the mobile or web through the user's smartphone, tablet PC or computer. Intelligent automatic feed for fish farms using artificial intelligence configured so that accurate and proper feeding can always be achieved through remote control even if the user is not at the farm site by being configured to access and remotely check and control the feeding process. It is to provide a supply system.

In order to achieve the above object, according to the present invention, an intelligent automatic feeding system for a fish farm using artificial intelligence includes a feeder configured to supply feed to the farm; and a remote control terminal for remotely controlling the feeder, wherein the feeder is configured to remotely supply feed according to a control command transmitted from the remote control terminal, and at the same time, the artificial intelligence (AI) ), an intelligent automatic feeding system for fish farms using artificial intelligence is provided, characterized in that it is configured to determine the optimal feeding amount for the farm and farmed fish and to automatically supply the feed according to the determined feeding amount. .

Here, the remote control terminal is characterized in that it is configured by installing a dedicated program for remote control and monitoring in a user's smartphone, tablet PC or computer (PC).

In addition, the feed supply unit includes a feed storage unit including a feed tank in which feed for supply to the farm is stored; a first slide opening and closing at least one side of the feed storage unit to discharge the stored feed; a first limit switch for controlling opening and closing of the first slide; a measuring unit including a measuring means for measuring the amount of feed discharged from the feed storage unit when the first slide is opened; a second slide opening and closing at least one side of the measuring unit; a second limit switch for controlling opening and closing of the second slide; at least one feed outlet for supplying the feed discharged from the measuring unit to the outside when the second slide is opened; a third slide for opening and closing the feed outlet; a blower unit including a blower for exporting the feed discharged from the measuring unit to the outside through the feed outlet; a communication unit comprising a wireless communication module for communication with an external device including the remote control terminal; A control panel provided on one side to input various control commands or set values for the supply amount and supply time of feed and the operation of the feed feeder; and controlling the overall operation of the feeder according to a control command input through the control panel or transmitted from the remote control terminal, and a fuzzy inference based Intelligent Fish Feeding Model; FIIFF reasoning model) to determine the optimal feeding amount for farms and farmed fish through artificial intelligence (AI), establish a supply plan according to the determined feeding amount, and automatically control feed supply. It is characterized in that it is composed of.

In addition, the feed feeder stores feed in the feed storage unit, and when the feed amount is input through the control panel or the remote control terminal provided on one side of the device during the feed supply, the feed amount is input according to the input value. 3 The slide moves in the feed feeding direction and moves until the first slide contacts the first limit switch to open the feed storage unit, thereby moving the feed stored in the feed storage unit to the measuring unit. and, when it is detected that the amount of feed as much as the input weight is accumulated in the measurement unit, the first slide is closed and the second slide moves until it contacts the second limit switch, thereby opening the measurement unit. Feed by weight is moved to the feed outlet side, and the feed moved to the feed outlet is exported to the outside while the blower of the blower operates according to the blowing strength input through the control panel or the remote control terminal. The feed is supplied, and when it is detected that all of the feed collected in the measuring unit has passed out, the second slide is moved to close the measuring unit and supply of the feed is terminated.

Furthermore, the feed feeder is characterized in that it is configured to simultaneously supply feed to a plurality of tanks by forming two or more feed outlets.

In addition, the feed feeder is configured to further include a protective separator for preventing backflow and damage caused by colliding feed with the blower of the blower unit, and adjusting the blowing strength of the blower unit step by step to discharge distance of the feed. And it is characterized in that it is configured to be able to adjust the position.

In addition, the feed feeder displays the remaining amount of feed in real time through the control panel and the remote control terminal installed on one side, and when the feed is insufficient, sends a notification to the user about feed shortage and supplementation, and operates the feed feeder. A monitoring function that sends a warning message informing the user of the corresponding content through the control panel and the remote control terminal when an abnormality occurs in the feed or feeding operation, and the control panel and the remote control terminal so that the desired amount of feed can be supplied at a desired time It is characterized in that it is configured to have a reservation supply function to set the feed supply time and supply amount through.

Furthermore, the FIIFF inference model may include a variable determination step of determining input variables and output variables of the FIIFF inference model; a fuzzy set determining step of determining a fuzzy set for input variables of the FIIFF inference model; a fuzzy rule generation step of generating fuzzy rules of the FIIFF inference model; a system construction step of constructing a FIIFF inference system using the fuzzy set and the fuzzy rule; And an evaluation and adjustment step of evaluating the performance of the built FIIFF inference system and making adjustments, the process of determining the final feeding amount in real time based on the Mamdani-style fuzzy inference system It is characterized in that it is configured to perform.

Here, in the variable determination step, from the data measured and collected for the growth environment of the farm, the growth of fish and the amount of feed supplied, the real-time temperature of the tank, the amount of dissolved oxygen (DO), the total weight of the fish species, The process of determining the input and output variables of the FIIFF inference model by setting the four factors including the swimming state of the fish as input variables and setting the final feed amount as an output variable is performed It is characterized in that it is configured to be.

In addition, the fuzzy set determining step is characterized in that, for each input variable determined in the variable determining step, a process of determining a fuzzy set is performed by dividing the range of each variable value by section according to a predetermined criterion. to be

In addition, the fuzzy rule generation step includes a process of generating a rule for setting a relationship between an input variable and an output variable by matching output variable values according to each input variable value of the fuzzy set determined in the fuzzy set determining step. It is characterized in that it is configured to perform.

Furthermore, in the system construction step, based on the fuzzy rules generated in the fuzzy rule generation step, the process of building the FIIFF inference system using MATLAB Fuzzy Logic Toolbox is configured to be performed. do.

In addition, in the evaluation and adjustment step, the input variables and output variables of the FIIFF inference system built in the system construction step are reviewed by comparing with the actual measured data of the farm, and the adjacent range of the fuzzy set is adjusted. Configured to be performed characterized by being

In addition, the intelligent automatic feeding system is configured to send and receive communication between a plurality of feeders and the remote control terminal, and analyzes the correlation between actual feed supply data and fish growth rate by fish species and farm Building and storing big data for automatic feeding, learning on the big data through artificial intelligence to build the FIIFF inference model, and determining and transmitting the optimal feed amount for each feeder Characterized in that it is configured to further include a server configured to perform.

Furthermore, according to the present invention, in the intelligent automatic feeding method for fish farms using artificial intelligence configured to supply feed to the farm using the intelligent automatic feeding system for fish farms using artificial intelligence described above, the intelligent Through the automatic feeding system, a fuzzy inference-based intelligent fish automatic feeding model (FIIFF inference model) configured to determine the final feeding amount in real time based on the Mamdani-style fuzzy inference system Feed amount determination step in which the optimal amount of feed is determined by using; and a feeding step in which feed is supplied through a feeder of the intelligent automatic feeding system according to the feeding amount determined in the feeding amount determining step. A method is provided.

As described above, according to the present invention, the process of accurately determining the appropriate feeding amount according to the fish species, growth, activity, etc. of farmed fish through artificial intelligence (AI) and supplying the necessary amount of feed evenly throughout the farm is automatically By providing an intelligent automatic feeding system for fish farms using artificial intelligence, which is configured to be performed, the farm manager arbitrarily feeds the farm manager depending on his or her experience without considering the size of the farm or the difference depending on the fish species and the feeding time. In addition to the failure to maintain the feeding amount and requiring a lot of labor, the feed is not evenly distributed throughout the farm due to manual spraying, which prevents the uniform growth of farmed fish. can be solved

In addition, according to the present invention, as described above, the process of accurately determining the appropriate feeding amount according to the fish species, growth, activity, etc. of farmed fish through artificial intelligence (AI) and supplying the necessary amount of feed evenly throughout the farm is automatic. By providing an intelligent automatic feeding system for fish farms using artificial intelligence configured to be performed, it is configured to supply feed only according to the time and amount of feeding simply predetermined by the user without accurate grounds based on the growth or activity of fish species. Inaccurate feed supply is intensively administered to only a part of the entire farm, which hinders the even growth of all fish in the farm, and damages the fish due to competition for food, as well as water pollution and disease due to oversupply of feed. It is possible to solve the problems of the automatic feeding machines of the prior art, which had problems causing

In addition, according to the present invention, as described above, the process of accurately determining the appropriate feeding amount according to the fish species, growth, activity, etc. of farmed fish through artificial intelligence (AI) and supplying the necessary amount of feed evenly throughout the farm is automatic. By being configured to be performed, it is possible to solve all the disadvantages of the existing manual feeding method and automatic feeding machines, and at the same time, various information on the overall feeding and aquaculture work obtained through artificial intelligence (AI) can be transferred to the user's smartphone. By providing an intelligent automatic feeding system for fish farms using artificial intelligence, which is configured to remotely check and control the feeding process by accessing via mobile or web via a tablet PC or computer, the user is not at the farm site. Through remote control, precise and appropriate feeding can always be achieved.

1 is a diagram schematically showing the overall configuration of an intelligent automatic feeding system for a fish farm using artificial intelligence according to an embodiment of the present invention.
2 is a diagram schematically showing the basic concept of an intelligent feeding control method applied to an intelligent automatic feeding system for a fish farm using artificial intelligence according to an embodiment of the present invention.
3 is an overall view of a Fuzzy Inference based Intelligent Fish Feeding Model (FIIFF inference model) applied to an intelligent automatic feeding system for a fish farm using artificial intelligence according to an embodiment of the present invention. This is a flowchart that outlines the process.
4 is a diagram showing fuzzy membership functions for four input variables generated based on a fuzzy set of input variables.
5 is a table showing fuzzy sets of input and output variables determined based on actual data measured in a grouper fish farm.
6 is a diagram showing some of 49 fuzzy set rules applied to an intelligent fish automatic feeding model (FIIFF inference model) based on fuzzy inference according to an embodiment of the present invention organized in a table according to fuzzy set notation.
7 is a diagram showing a process of setting a membership function using a membership function editor for input variables for a performance evaluation experiment of a FIIFF inference model according to an embodiment of the present invention.
8 is a diagram illustrating a process of generating fuzzy rules using a fuzzy rule editor for a performance evaluation experiment of a FIIFF inference model according to an embodiment of the present invention.
9 is a diagram showing a process of displaying the result of calculating the final feeding amount through a performance evaluation experiment of the FIIFF inference system according to an embodiment of the present invention.
10 is a graph showing the result of comparing the actual feeding amount of the fish farm tank with the feeding amount calculated through the FIIFF inference model.
11 is a result of comparing the average monthly feed amount actually fed to measure the overall efficiency of feed feed and the feed amount measured monthly by the FIIIFF fuzzy reasoning-based intelligent fish automatic feeding model according to an embodiment of the present invention It is a diagram showing the arrangement in a table.

Hereinafter, specific embodiments of an intelligent automatic feeding system for a fish farm using artificial intelligence according to the present invention will be described with reference to the accompanying drawings.

Here, it should be noted that the contents described below are only one embodiment for carrying out the present invention, and the present invention is not limited to the contents of the embodiments described below.

In addition, in the following description of the embodiments of the present invention, for parts that are the same as or similar to the contents of the prior art or are determined to be easily understood and implemented at the level of those skilled in the art, the detailed descriptions are provided to simplify the description. It should be noted that .

Subsequently, with reference to the drawings, specific details of the intelligent automatic feeding system for fish farms using artificial intelligence according to the present invention will be described.

First, referring to FIG. 1, FIG. 1 is a diagram schematically showing the overall configuration of an intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention.

As shown in FIG. 1, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention is largely divided into a feeder 20 for supplying feed to the farm, and a feeder ( 20) may be configured to include a remote control terminal 30 for remotely controlling.

Here, the above feed feeder 20, as shown in FIG. 1, includes a feed storage unit 21 including a feed tank for storing feed for supplying to the farm, and feed stored in the feed storage unit 21 A first slide 22 for opening and closing at least one side of the feed storage unit 21 to discharge the food, a first limit switch 23 for controlling the opening and closing of the first slide 22, and a first slide 22 ), a measuring unit 24 comprising a measuring means for measuring the amount of feed discharged from the feed storage unit 21 when the feed storage unit 21 is opened, and a second slide 25 for opening and closing at least one side of the measuring unit 24; A second limit switch 26 for controlling the opening and closing of the second slide 25, and at least one feed outlet for supplying the feed discharged from the measuring unit 24 to the outside when the second slide 25 is opened ( 27), a third slide 28 for opening and closing the feed outlet 27, and a blower unit including a high-pressure blower for exporting the feed discharged from the measuring unit 24 to the outside through the feed outlet 27 ( 29) and although not shown, a communication unit including a wireless communication module for communication with an external device such as the remote control terminal 30 and a control panel for setting the operation of the device, such as feed amount or supply time. And, according to the control command input through the control panel or transmitted from the remote control terminal 30, or, as described later, through artificial intelligence (AI), including a control unit configured to automatically control the operation of the entire device can be configured.

More specifically, the basic operation of the above feed feeder 20 is to store feed in the feed storage unit 21 and supply feed through a control panel or remote control terminal 30 provided on one side of the device. When the feeding amount is input, the third slide 28 moves in the feed feeding direction according to the input value, and the first slide 22 moves until it contacts the first limit switch 23 to feed storage unit ( 21) moves the feed stored in the feed storage unit 21 to the measurement unit 24.

Next, when it is detected that the weight of feed accumulated in the measuring unit 24 is detected, the first slide 22 is closed and the second slide 25 moves until it contacts the second limit switch 26 By opening the measuring unit 24, feed as much as the input weight is sent to the side of the feed outlet 27.

Subsequently, feed is supplied by sending the feed to the outside through the feed outlet 27 while the blower of the blower 29 is operated according to the blowing intensity input through the control panel or the remote control terminal 30. , When it is detected that all of the feed collected in the measuring unit 24 has escaped, the second slide 25 is moved to close the measuring unit 24 and feed feeding is terminated. Therefore, through this configuration, accurate weight measurement per fish By calculating the required amount of feed, the correct amount of feed can be supplied.

Here, since a more specific configuration or operation of the feed feeder 20 as described above can be appropriately configured by a person skilled in the art with reference to the configuration of prior art feed supply devices, in the present invention, in order to simplify the description, the prior art It should be noted that the detailed description has been omitted for contents that can be easily understood and implemented by those skilled in the art by referring to the literature of.

More specifically, the feeder 20 according to the embodiment of the present invention is configured to prevent rust by applying a plastic feed container and a double hopper measuring box, for example, in consideration of the characteristics to be installed near the aquaculture tank. It can be configured to improve the disadvantages of the existing feed supply device that can supply feed to only one tank by forming two or more feed outlets 27 to supply feed to a plurality of tanks at the same time. there is.

Furthermore, in the feed feeder 20 according to the embodiment of the present invention, a protective separator for preventing backflow and damage caused by rapid encounter between air and feed is installed in the blowing device of the blowing unit 29, and the blowing strength is measured by code. By adjusting the value, for example, by making it possible to adjust the feed discharge distance in three stages such as strong (6M), medium (4M), and weak (2M), it can be configured to enable feed discharge to various positions. there is.

In addition, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention, as shown in FIG. 1, further includes a remote control terminal 30 to enable remote control and monitoring. can be configured.

Here, the remote control terminal 30, preferably, as shown in FIG. 1, for example, installs a dedicated program for remote control and monitoring on a user's smartphone, tablet PC or computer (PC). Although it can be easily implemented without the need to configure separate hardware by installing it, it is configured to exchange various control commands or set values such as feed amount or time through wired or wireless communication with the feed feeder 20 If so, there is no particular limitation.

That is, the feed feeder 20 according to the embodiment of the present invention displays the remaining amount of feed in real time through a control panel installed on one side or a remote control terminal 30 connected to the feed feeder 20 through wired or wireless communication. When the feed is insufficient, a notification is sent to the user about feed shortage and supplementation, and when an abnormality occurs in the operation of the feed feeder 20 or the feed supply operation, a warning notifying the user of the corresponding content through the control panel and the remote control terminal 30 A monitoring function that sends a message, and reservation supply that sets the feed supply time in minutes and sets the supply amount in grams (g) through the control panel or remote control terminal 30 to supply the desired amount of feed at the desired time It can be configured to provide a function, whereby it can be configured to enable accurate feeding anytime, anywhere in conjunction with a PC and mobile.

In addition, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention is configured to enable remote feeding in conjunction with a PC and mobile as described above, in addition to artificial intelligence Through this, it can be configured to establish an optimal supply plan for the fish and automatically supply feed.

Subsequently, the specific contents of the artificial intelligence algorithm applied to the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention will be described.

That is, in the embodiment of the present invention, in order to solve the problems of the existing automatic feeding device and to maximize the efficiency of feeding while properly maintaining the growth rate of farmed fish in a land-based fish farm, fuzzy reasoning-based presented a Fuzzy Inference based Intelligent Fish Feeding Model (FIIFF inference model), through which the growth environment information and growth activity status of farmed fish, which were not considered by existing automatic feeding devices, can be obtained in real time. It can automatically provide feed supply by measuring and calculating in real time the most suitable feed supply amount for the current growth status of farmed fish.

Here, in the embodiments of the present invention described below, the present invention has been described by taking the case of grouper fish culture as an example to which the present invention can be applied, but the present invention is not necessarily limited to the case described in the following examples, That is, it should be noted that the present invention is applicable to various types of fish farming in addition to grouper farming, and can be applied by appropriately changing and modifying it as needed by those skilled in the art.

In more detail, for example, grouper fish farming in Korea is carried out in three ways: land tank type, sea cage type, and festival type, but most of them are land tank type. , the appropriate temperature is 15℃ ~ 19℃, the amount of dissolved oxygen is 6 ~ 7mg/ℓ, the formulated feed is floating feed, and the size of the feed is efficient to supply large-sized feed within the ingestible range. The supply amount depends on the breeding water temperature and grouper size.

In addition, at the current grouper fish farming site, feed is supplied very close to fullness (full stomach), but the feed efficiency decreases as the amount of feed is more than the proper amount and closer to full stomach. Therefore, if you save about 10% of the feed and see the same growth effect as full feeding, you can effectively reduce the production cost of aquaculture considering the proportion of feed in the production cost of aquaculture.

In most domestic aquaculture farms, managers with a lot of experience in aquaculture manually supply the amount of feed or use an automatic feeder to save labor costs. Setting should be preceded, which should vary depending on water temperature, body size, weight, calorie and moisture content in feed, etc. The standard amount of aquaculture feed by experience is described in prior art literature, but there are many problems in actual application.

In addition, since the general method for determining the feeding amount supplied by the farm is very ambiguous depending on the water quality environment such as the current temperature of the farm or the amount of dissolved oxygen and the experience of the aquaculture manager, in the present invention, the ambiguous feeding amount is determined. In order to improve the process more rationally, based on fuzzy logic, the water quality environment information and fish activity status are expressed as a fuzzy set, and the farm manager's experience is generated as a fuzzy rule to determine the final feeding amount using a fuzzy inference method. configured to produce

To this end, the present invention, as described later, an intelligent fish automatic feeding model based on fuzzy inference configured to determine the final feeding amount in real time based on a Mamdani-style fuzzy inference system ( FIIFF inference model) was presented.

In other words, the Mamdani-type fuzzy reasoning method is a fuzzy reasoning method first proposed by London University professor Ebrahim Mamdani in 1975 to control a boiler-coupled steam engine, and is one of the representative reasoning methods of fuzzy reasoning. is largely divided into four stages, including fuzzification of input variables, rule evaluation, and integration and defuzzification of output rules.

Here, since more specific details of the Mamdani-type fuzzy inference technique are obvious to those skilled in the art through prior art documents, etc., therefore, in the present invention, in order to simplify the description, as described above, through prior art documents, etc. It should be noted that detailed descriptions of contents that can be easily understood and implemented by those skilled in the art have been omitted.

Continuing to refer to FIG. 2, FIG. 2 is a diagram schematically showing the basic concept of an intelligent feeding control method applied to the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention. am.

As shown in FIG. 2, the intelligent feeding control method presented in the present invention is a method of controlling a feeding device by measuring the feeding desire of fish, and using the video or image of the movement degree and movement pattern of the fish to determine the feeding desire. It may be configured to measure, determine the optimal supply amount by applying the calculated water quality environment map based on water quality environment information such as the amount of dissolved oxygen and water temperature, and establish a supply plan accordingly.

In addition, as a fish behavior state tracking model for measuring feeding desire, each fish is individually assigned an ID in the image data for learning fish behavior to recognize each fish object, and the size, up-down trajectory and swimming for each fish object ID By measuring speed, it correlates fish behavior and conditions, e.g., stationary (very full), slow swimming (normally full), fast swimming (very hungry), abnormal swimming (disease, stress), etc. It can be configured to be classified into four types.

In addition, as an artificial intelligence model for deriving feed rate, feed amount with a total of four input variables including temperature, dissolved oxygen (DO), individual weight, and activity state (state) ) as an output variable, an intelligent automatic feeding system for automatic feeding amount control is implemented.

That is, as described above, by using the big data related to the fish growth environment and monitoring information on the fish growth activity status, for example, the feed supply of fish including one input layer, one output layer, and three hidden layers An intelligent fuzzy inference engine configured to infer the feeding amount automatically can be built.

More specifically, referring to FIG. 3, FIG. 3 is a fuzzy reasoning-based intelligent fish automatic feeding model applied to the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention ( This is a flowchart schematically showing the overall process of the FIIFF inference model).

As shown in FIG. 3, the process of determining the optimal feed supply amount using a fuzzy reasoning-based intelligent fish automatic feeding model (FIIFF inference model) according to an embodiment of the present invention is divided into broad categories, first, FIIFF inference A variable determination step (S10) for determining the input and output variables of the model, a fuzzy set determination step (S20) for determining a fuzzy set for the input variables of the FIIFF inference model, and a fuzzy rule for generating a fuzzy rule of the FIIFF inference model Rule generation step (S30), system construction step (S40) of building a FIIFF inference system using the determined fuzzy set and fuzzy rules, and evaluation and adjustment step of evaluating and adjusting the performance of the built FIIFF inference system (S50 ).

Here, the variable determination step (S10) is a step of determining the input variables and output variables of the FIIFF inference model, for example, as described above, the real-time temperature of the fish farm tank, the amount of dissolved oxygen (DO) It may be configured to set four factors, such as , the weight of the individual fish species, and the swimming state of the fish, as input variables, and set the output variable as the final feeding amount.

In addition, in this embodiment, the final feeding amount was calculated through the FIIFF inference model by using the data collected from the Goheung Branch of the Goheung Branch of the Jeollanam-do Institute of Maritime Science and Fisheries Science located in Goheung, Jeollanam-do as an input variable. To this end, For the grouper fish species, measure the average weight change of fish every day for 2,000 grouper fish in one tank for 8 months from April to November 2019, and input water temperature, dissolved oxygen amount and fish activity status data according to the average total weight of the fish The final output variable was determined by the activity status of fish according to the weight change of grouper fish, the water temperature of the farm, and the feeding amount of fish related to the amount of dissolved oxygen.

In addition, referring to FIGS. 4 and 5, FIG. 4 is a diagram showing fuzzy membership functions for four input variables generated based on a fuzzy set of input variables, and FIG. 5 is based on actual data measured in group fish farms. It is a diagram showing fuzzy sets of determined input and output variables organized in a table.

Here, in the table shown in FIG. 5, BL/sec represents the swimming speed of fish as Body/sec, and the feeding amount (F) represents the total feeding amount of one tank for 2,000 individuals.

As shown in FIGS. 4 and 5, the final feeding amount of the farm must be appropriately adjusted according to the value of the input variable, and the fuzzy set of these input variables can be determined by referring to the grouper standard manual and the actual data of the farm. .

That is, as shown in FIGS. 4 and 5, the fuzzy set determination step (S20) is performed for each input variable, such as High, Normal, and Low. It may be configured to perform a process of determining a fuzzy set by dividing the range of each variable value by section according to a predetermined criterion.

Next, referring to the fuzzy rule generation step (S30), the fuzzy rule in the farm can be determined according to the opinions of aquaculture experts with years of experience. Feeding rules were created based on the knowledge experienced while operating the farm for two years.

More specifically, in the above embodiment, since there are 4 input variables (temperature, dissolved oxygen amount, weight of animal, activity state) and 1 output variable (feeding amount), up to 64 fuzzy rules can be generated. In the embodiment, 49 rules were created according to the weight, water temperature, dissolved oxygen content, and activity status of the fish species that have the greatest influence on the feeding amount in the current state of the farm.

That is, referring to FIG. 6, FIG. 6 shows some of the 49 fuzzy set rules applied to the intelligent fish automatic feeding model (FIIFF inference model) based on fuzzy reasoning according to an embodiment of the present invention in a table according to the fuzzy set notation. It is a schematic representation.

As shown in FIG. 6, for example, when the object weight (W) and temperature (T) are both LOW, and the amount of dissolved oxygen (DO) and activity state are NONE, the value of the output variable (F) becomes VERY_LOW, and the object When the weight (W) is HIGH and the temperature (T), dissolved oxygen content (DO), and activity state are all NONE, the value of the output variable (F) becomes HIGH.

Therefore, as shown in FIG. 6, in the fuzzy rule generating step (S30), according to each input variable value of the fuzzy set determined in the fuzzy set determining step (S20) as described above with reference to FIGS. 4 and 5 It may be configured so that a process of generating a rule for establishing a relationship between an input variable and an output variable by matching values of the output variable with each other is performed.

Next, in the system construction step (S40), a process for building a fuzzy inference system based on the 49 fuzzy rules generated as described above may be performed. For this purpose, for example, MATLAB Using a commercial program such as Fuzzy Logic Toolbox, an intelligent automatic feeding fuzzy inference system can be constructed using 49 fuzzy rules according to the opinions of farm experts.

Therefore, it is possible to build a FIIFF inference system as described above, and it is necessary to evaluate and adjust the built inference system in order to calculate a more accurate final feeding amount.

That is, in the above evaluation and adjustment step (S50), the input and output variables of the FIIFF inference model constructed as described above are compared with the measured data related to the actual fish growth and feed supply in the farm, and the input variables are reviewed. A process of adjusting the contiguous range of the fuzzy set may be performed by fully considering the correlation of .

Here, the most important input variable factor is the observation of the swimming state of the fish according to the temperature and dissolved oxygen amount of the current fish farm during the fish farm. The fuzzy rule of the FIIFF inference system was adjusted, and the defuzzification method for calculating the final feeding amount used the center of gravity method to provide consistent results.

Subsequently, in order to evaluate the performance of the FIIFF inference model according to the embodiment of the present invention implemented as described above, the FIIFF inference system to which the FIIFF inference model according to the embodiment of the present invention is applied is actually implemented to conduct performance experiments and evaluations. Describe the results of the process.

That is, the present inventors use Windows 10 Pro as an operating system, MATLAB Fuzzy Logic ToolBox V2.0 as an implementation tool, and MATLAB as an implementation language in a hardware environment using Intel i7-7820X CPU, 32GB RAM, and TITAN Xp GPU. The system was implemented, and performance tests were conducted using this FIIFF inference system.

More specifically, referring to FIG. 7, FIG. 7 is a membership function editor for four input variables (whole weight, temperature, dissolved oxygen content, and activity state) for a performance evaluation experiment of the FIIFF inference system according to an embodiment of the present invention. It is a diagram showing the process of setting membership functions for four input variables using

Also, referring to FIG. 8, FIG. 8 is a diagram showing a process of generating 49 fuzzy rules using a fuzzy rule editor for a performance evaluation experiment of the FIIFF inference system according to an embodiment of the present invention.

In addition, referring to FIG. 9, FIG. 9 is a diagram showing a process in which the result of calculating the final feeding amount through the performance evaluation experiment of the FIIFF inference system according to an embodiment of the present invention is displayed.

Here, the calculation result shown in FIG. 9 is the final calculation when the total weight of one individual is 57.0 g, the water tank temperature is 23.3 degrees, the amount of dissolved oxygen is 1.44 mg / ℓ, and the activity state is slowly swimming (0.31). It shows that the total feeding amount of one tank is 121Kg.

That is, the inventors calculated using the actual feeding data and the FIIFF inference system according to the embodiment of the present invention based on the water quality environment data and feeding amount measured from April to November 2019 in the actual grouper fish farm The performance evaluation experiment was conducted by comparing the recorded data. At this time, the actual data measured at a specific time through the water temperature sensor and DO sensor installed in the actual grouper fish farm were used for the water temperature and dissolved oxygen amount, and the fish activity status was measured using a video camera. Therefore, grouper swimming speed was calculated using video data taken just before morning feeding time, and hunger and fullness according to swimming speed were converted into values between [0.0 and 1.0] and used in the experiment.

In addition, the average weight (total weight) of grouper fish used in this experiment was 30.7g in April when it was first stocked, and as a result of continuous growth by eating feed every day, the average weight in November grew to 219.6g. In the embodiment, the average total weight of such an individual and the number of individuals in one tank were assumed to be 2,000, and the results of calculating the daily feeding amount through the FIIFF inference model were compared.

More specifically, referring to FIG. 10, FIG. 10 shows the actual feeding amount of the farm tank based on the total weight of grouper fish, farm temperature, dissolved oxygen amount, and swimming speed from April 19 to August 5, 2019 It is a graph showing the result of comparing the feeding amount calculated through the FIIFF inference model.

Here, the graph shown in FIG. 10 shows the actual feeding amount for 3 days per month and the calculation result of the FIIFF inference system due to the inability to display all of the daily measurement values and calculation results for 8 months due to various conditions.

In addition, in FIG. 10, the actual feeding amount was provided the same for a certain period of each month according to the growth of the individual, and the result calculated by the FIIFF inference system according to an embodiment of the present invention is relatively less than the actual feeding amount indicates that it is produced.

Continuing, referring to FIG. 11, FIG. 11 shows the average monthly feed amount actually fed based on the average total weight of the grouper fish farm, the temperature of the tank, the amount of dissolved oxygen, and the activity status of the fish in order to measure the overall efficiency of the fed feed. It is a diagram showing the results of comparing the feeding amounts measured monthly with the FIIIFF fuzzy reasoning-based intelligent fish automatic feeding model according to an embodiment of the invention in a table.

As shown in the experimental results of FIGS. 10 and 11, using the FIIFF inference system according to an embodiment of the present invention, the monthly average of 37.05Kg when the daily feeding amount is adjusted according to the fish weight, temperature, dissolved oxygen amount, and swimming state It can be seen that the feeding amount can be reduced, and the total feeding amount for 8 months can be reduced by 14.8% compared to the actual feeding amount.

That is, as described above, using the FIIFF inference system, which is an intelligent fish automatic feeding model based on fuzzy reasoning according to an embodiment of the present invention, the current real-time water quality environment of the farm and the activity status of farmed fish are measured, and the degree of water quality and By modeling the degree of hunger with a fuzzy membership function to calculate the final feeding amount, the farm manager does not consider the current hunger and fullness status at all, and only supplies a certain amount of feed at a certain time based on the experience of the farm manager. Therefore, it is possible to solve the disadvantages of the existing automatic feeding model and automatic feeding devices for fish farms, which had limitations in not being able to adjust the amount of feed according to the activity status of the fish currently being farmed.

In addition, by collecting multi-year data on actual feed supply for each fish species and farm, and analyzing the correlation with the fish growth rate based on the supplied feed amount to build big data for automatic feeding, The accuracy of the final feeding amount calculation result of the intelligent automatic feeding FIIFF inference model according to the embodiment can be further increased.

Furthermore, although not shown, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention is a server that communicates with and receives communication between the feeder 20 and the remote control terminal 30. It may be configured to further include, and through this configuration, it is possible to easily control a plurality of feeders 20 installed in different locations remotely.

In addition, the intelligent automatic feeding system 10 for fish farms using artificial intelligence according to an embodiment of the present invention, as described above, has a correlation between multi-year data on actual feed supply by fish species and farms and fish growth rate. Analyzing, big data for automatic feeding is built and stored in the server, and based on this big data, the server performs learning on the big data to build an inference system such as the FIIFF inference model described above, and each By being configured to determine and transmit the optimal feed supply amount for the feeder 20, since the operation of supplying feed with the optimal supply amount to the plurality of feeders 20 can be automatically performed, each feeder 20 It is possible to further simplify the configuration of, and at the same time increase the efficiency and productivity of fish farming.

Therefore, as described above, an intelligent automatic feeding system for a fish farm using artificial intelligence according to an embodiment of the present invention can be easily implemented.

That is, as described above, in the embodiment of the present invention, in order to solve the problems of the existing automatic feeding device and to maximize the efficiency of feed supply while properly maintaining the growth rate of farmed fish in a land-based fish farm, fuzzy reasoning-based presented an intelligent fish automatic feeding model (FIIFF inference model), and through this, the intelligent automatic feeding system 10 for fish farms using artificial intelligence according to an embodiment of the present invention, existing automatic feeding devices It may be configured to measure the growth environment information and growth activity status of farmed fish that have not been considered in real time, calculate the feed supply amount most suitable for the current growth state of the farmed fish in real time, and automatically provide the feed supply amount.

To this end, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence to which the FIIFF model presented in the embodiment of the present invention is applied measures the temperature and dissolved oxygen amount of the fish tank in real time, as described above, and According to the activity (swimming) state according to the body weight (whole weight), the fish is measured by dividing it into very hungry, moderately hungry, and full. Then, water temperature and dissolved oxygen (DO) are input variables for fuzzy inference. , Fish Weight, and Fish Swimming State data may be input, and processing of calculating the final feeding amount by applying the Mamdani-type fuzzy reasoning rule may be performed.

In addition, the FIIFF inference model presented in the embodiment of the present invention, as described above, was 14.8 times higher than the actual feeding amount in the farm for 8 months from the results of the feeding amount calculation experiment of the FIIFF inference model conducted by the present inventors. % reduction effect, and therefore, the intelligent automatic feeding system 10 for fish farms using artificial intelligence to which the FIIFF inference model presented in the embodiment of the present invention is applied is the current growth environment information and real-time activity status of farmed fish Since the feeding amount is calculated based on the feed amount, the appropriateness of the feeding amount is very high, and it will improve the fundamental problems of the existing automatic feeding system for fish to develop a very underdeveloped aquaculture facility into a more advanced and intelligent smart farm. hope to be able to

Moreover, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention, as described above, has a time reservation feeding function in units of minutes, which was not found in existing automatic feeding feeders, and has two outlets. Formed to supply feed to two tanks at the same time, it is possible to improve the disadvantages of the existing feed supply device that was able to supply feed to only one tank, and air and feed rapidly collide with the blower device of the feed discharge pipe. By installing a protective membrane to prevent backflow and breaking, and by adjusting the blower strength to a code value, the discharge distance is adjusted in three stages: strong (6M), medium (4M), and weak (2M), so that feed is discharged to various locations. It can be configured to be possible.

In addition, the intelligent automatic feeding system 10 for a fish farm using artificial intelligence according to an embodiment of the present invention uses AI (artificial intelligence) to enable precise and more precise feeding of feed. In this way, by combining the artificial intelligence function, the volume of the device or the size of the control part is reduced compared to the automatic feeding machine of the prior art, enabling various space utilization and facilitating movement and installation. It is possible to supply feed remotely by accessing the mobile phone or web even when not in the hospital, so it has the advantage of being able to supply feed remotely without being tied to the site.

In the above, the details of the intelligent automatic feeding system for fish farms using artificial intelligence according to the present invention have been described through the embodiments of the present invention as described above, but the present invention is limited to the contents described in the above embodiments. Therefore, it is natural that the present invention is capable of various modifications, changes, combinations, and substitutions according to design needs and other various factors by those skilled in the art to which the present invention belongs. would.

10. Intelligent automatic feeding system for fish farms using artificial intelligence
20. Feed feeder 21. Feed storage unit
22. 1st slide 23. 1st limit switch
24. Measuring part 25. Second slide
26. 2nd limit switch 27. Feed outlet
28. Slide 3 29. Blower
30. Remote control terminal

Claims (15)

In the intelligent automatic feeding system for fish farms using artificial intelligence,
A feeder configured to supply feed to the farm; and
It is configured to include a remote control terminal for remotely controlling the feeder,
The feed feeder,
A feed storage unit including a feed tank in which feed for supply to the farm is stored;
a first slide opening and closing at least one side of the feed storage unit to discharge the stored feed;
a first limit switch for controlling opening and closing of the first slide;
a measuring unit including a measuring means for measuring the amount of feed discharged from the feed storage unit when the first slide is opened;
a second slide opening and closing at least one side of the measuring unit;
a second limit switch for controlling opening and closing of the second slide;
at least one feed outlet for supplying the feed discharged from the measuring unit to the outside when the second slide is opened;
a third slide for opening and closing the feed outlet;
a blower unit including a blower for exporting the feed discharged from the measuring unit to the outside through the feed outlet;
a communication unit comprising a wireless communication module for communication with an external device including the remote control terminal;
A control panel provided on one side to input various control commands or set values for the supply amount and supply time of feed and the operation of the feed feeder; and
The overall operation of the feeder is controlled according to a control command input through the control panel or transmitted from the remote control terminal, and a fuzzy inference based Intelligent Fish Feeding Model (FIIFF inference model) ), including a control unit that determines the optimal feeding amount for farms and farmed fish through artificial intelligence (AI), establishes a supply plan according to the determined feeding amount, and automatically controls feed supply By being composed
It is configured to remotely supply feed according to the control command transmitted from the remote control terminal, and at the same time, by using artificial intelligence (AI), the optimal feeding amount for the farm and farmed fish is determined, and feed supply is performed according to the determined feeding amount. An intelligent automatic feeding system for fish farms using artificial intelligence, characterized in that it is configured to be automatic.
According to claim 1,
The remote control terminal,
An intelligent automatic feeding system for fish farms using artificial intelligence, characterized in that it is configured by installing a dedicated program for remote control and monitoring on a user's smartphone, tablet PC or computer (PC).
delete According to claim 1,
The feed feeder,
When feed is stored in the feed storage unit and feed amount is input through the control panel or the remote control terminal provided on one side of the device during feed supply, the third slide moves in the feed feeding direction according to the input value. and moving the feed stored in the feed storage unit to the measurement unit by moving until the first slide contacts the first limit switch and opening the feed storage unit,
When it is sensed that feed as much as the input weight is accumulated in the measurement unit, the first slide is closed and the second slide is moved until it contacts the second limit switch, thereby opening the measurement unit as much as the input weight. Supply of feed by moving the feed to the feed outlet side and sending the feed to the feed outlet to the outside while the blower of the blower operates according to the blowing intensity input through the control panel or the remote control terminal this is done,
An intelligent automatic feeding system for a fish farm using artificial intelligence, characterized in that configured to move the second slide to close the measurement unit and terminate the supply of feed when it is detected that all the feed collected in the measurement unit has escaped.
According to claim 4,
The feed feeder,
An intelligent automatic feeding system for fish farms using artificial intelligence, characterized in that the feed outlet is formed in two or more to simultaneously supply feed to a plurality of tanks.
According to claim 5,
The feed feeder,
It is configured to further include a protective separator for preventing backflow and damage caused by collision of the feed with the blower of the blowing unit,
An intelligent automatic feeding system for fish farms using artificial intelligence, characterized in that configured to adjust the feeding distance and position of the feed by adjusting the blowing strength of the blowing unit step by step.
According to claim 6,
The feed feeder,
Displays the remaining amount of feed in real time through the control panel and the remote control terminal installed on one side, sends a notification to the user about lack of feed and replenishment when the feed is insufficient, and when an abnormality occurs in the operation of the feeder or the feed supply operation A monitoring function that sends a warning message informing the user of the corresponding content through the control panel and the remote control terminal, and
Intelligent for fish farms using artificial intelligence, characterized in that it is configured to have a reservation supply function for setting feed supply time and supply amount through the control panel and the remote control terminal so that a desired amount of feed can be supplied at a desired time. Automatic feeding system.
According to claim 1,
The FIIFF inference model,
a variable determination step of determining input variables and output variables of the FIIFF inference model;
a fuzzy set determining step of determining a fuzzy set for input variables of the FIIFF inference model;
a fuzzy rule generation step of generating fuzzy rules of the FIIFF inference model;
a system construction step of constructing a FIIFF inference system using the fuzzy set and the fuzzy rule; and
Process for determining the final feeding amount in real time based on the Mamdani-style fuzzy inference system, including evaluation and adjustment steps of evaluating the performance of the built FIIFF inference system and making adjustments An intelligent automatic feeding system for fish farms using artificial intelligence, characterized in that configured to be.
According to claim 8,
In the variable determination step,
Real-time temperature of the tank (Temperature), dissolved oxygen (DO), total weight of fish species, swimming state of fish from data collected and measured on the growth environment of the farm, fish growth and feed supply Four elements including a are set as input variables, and the final feeding amount (Feed) is set as an output variable to perform processing for determining the input variables and output variables of the FIIFF inference model. Artificial, characterized in that Intelligent automatic feeding system for fish farms using intelligence.
According to claim 9,
The fuzzy set determination step,
For fish farms using artificial intelligence, characterized in that for each input variable determined in the variable determination step, a process of determining a fuzzy set is performed by dividing the range of each variable value by section according to a predetermined criterion. Intelligent automatic feeding system.
According to claim 10,
The fuzzy rule generation step,
Artificial intelligence characterized in that it is configured to perform a process of generating a rule for setting a relationship between an input variable and an output variable by matching the value of an output variable according to each input variable value of the fuzzy set determined in the fuzzy set determining step to each other. Intelligent automatic feeding system for fish farms using intelligence.
According to claim 11,
The system construction step is,
Based on the fuzzy rules generated in the fuzzy rule generation step, processing for constructing the FIIFF inference system using MATLAB Fuzzy Logic Toolbox is performed. For fish farms using artificial intelligence, characterized in that Intelligent automatic feeding system.
According to claim 12,
In the evaluation and adjustment step,
The input variables and output variables of the FIIFF inference system built in the system construction step are compared with the actual measured data of the farm, reviewed, and the processing of adjusting the contiguous range of the fuzzy set is performed. Using artificial intelligence, characterized in that configured to perform Intelligent automatic feeding system for fish farms.
According to claim 13,
The intelligent automatic feeding system,
It is made to send and receive communication between the plurality of feeders and the remote control terminal, and analyzes the correlation between actual feed supply data and fish growth rate by fish species and farm to build big data for automatic feeding storage, and by performing learning on the big data through artificial intelligence to build the FIIFF inference model, and determining and transmitting the optimal feed supply amount for each feeder. A server configured to perform processing configured to further include An intelligent automatic feeding system for fish farms using artificial intelligence, characterized in that.
Fish farming using artificial intelligence configured to supply feed to the farm using the intelligent automatic feeding system for fish farming using artificial intelligence according to any one of claims 1, 2, and 4 to 14 In the enteric intelligent automatic feeding method,
Through the intelligent automatic feeding system, a fuzzy inference-based intelligent fish automatic feeding model (FIIFF inference model configured to determine the final feeding amount in real time based on a Mamdani-style fuzzy inference system) ) Determine the amount of benefits to determine the optimal amount of benefits using; and
An intelligent automatic feeding method for a fish farm using artificial intelligence, characterized in that it comprises a feeding step in which feed is supplied through a feeder of the intelligent automatic feeding system according to the feeding amount determined in the feeding amount determining step. .

Images