Method for productivity management for land-based fish farm through data prediction based on growth period

Abstract

The present invention relates to a method for managing fish productivity in land-based aquaculture farms through data prediction by growth stage. A method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage according to the present invention comprises: (a) a step of predicting the amount of feed input into a tank using tank sensing data; and (b) a step of predicting fish growth using the result of the prediction of the amount of feed input.

Description

Technology Field

[0001] The present invention relates to a method for managing fish productivity in land-based aquaculture farms through data prediction by growth stage. Background Technology

[0003] According to conventional technology, most land-based aquaculture farms are operated based on human empirical data, which presents limitations in farm automation and challenges in transferring aquaculture technology. There is a need to create a module that models and predicts fish growth; through such a fish growth model, it is possible to maximize operational efficiency, resolve the issue of technology transfer, and expand the land-based aquaculture market.

[0004] For the smooth operation of the fish growth prediction module, it is necessary to collect accurate data on fish farm growth. While using automatic feed dispensers allows for precise determination of feed quantities and reduces human resources, the high installation costs make it difficult for general fish farms to install such equipment. Furthermore, since there are many cases where feed forms that are difficult to administer mechanically, such as paste feed, are fed during rearing, the accurate measurement of feed input quantities often requires manual entry by humans. In actual fish farms, human error can lead to omissions regarding manually entered data such as feed quantities, and errors or gaps in this manual data result in a decrease in the accuracy of the fish farming model.

[0005] In addition, in land-based aquaculture farms, fish are periodically sorted based on size during growth and placed into a single tank with fish of similar size and weight. However, there are limitations in this process, such as the difficulty in collecting data on fish growth and the difficulty in predicting fish growth throughout the entire period using growth models. The problem to be solved

[0007] The present invention is proposed to solve the aforementioned problems and aims to provide a method for managing fish productivity in land-based aquaculture farms through data prediction by growth stage, which enables improving the productivity of the aquaculture farm and increasing operational efficiency by using a learned input feed amount prediction module and a segmented fish growth prediction module. means of solving the problem

[0009] A method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage according to the present invention comprises: (a) a step of predicting the amount of feed input into a tank using tank sensing data; and (b) a step of predicting fish growth using the result of the prediction of the amount of feed input.

[0010] Step (a) above performs a prediction of the amount of feed input using the result of a function calculation using fish body weight, oxygen content, and water temperature.

[0011] Step (a) above performs a prediction of the amount of feed input using a lookup table for the amount of oxygen consumed and the amount of feed supplied.

[0012] Step (b) above performs a prediction of the fish growth by operating growth prediction models divided into weight units in parallel.

[0013] Step (b) above performs a prediction of the sorting time using the weight value prediction result and the weight dispersion value.

[0014] A fish productivity management device for a land-based aquaculture farm according to the present invention comprises an input unit that receives input information including tank information and environmental information, a memory that stores a program for predicting the amount of feed input using the input information, and a processor that executes the program, wherein the processor divides a model for predicting growth by selection time based on the prediction results of the amount of feed input, and outputs the prediction results for the fish growth rate and selection time. Effects of the invention

[0016] According to the present invention, a model is presented that automatically predicts the amount of feed input for a fish growth prediction model using sensor data from a land-based aquaculture farm, thereby making it possible to reduce human error caused by manual input of the input feed amount. Furthermore, by predicting growth gaps that occur during fish sorting, a fish growth prediction module is proposed that adaptively operates on the fish sorting operations that occur periodically in land-based aquaculture farms through a learning model.

[0017] According to the present invention, when sorting occurs, a fish model for the next period is applied to provide a more accurate fish growth model by predicting the weight of the fish at each sorting time in the tank, rather than predicting the weight of the fish for the entire cycle in the tank.

[0018] According to the present invention, the accuracy of the growth model can be improved, and the efficiency of fish rearing can be increased, which can have a positive effect on the operation of land-based fish farms.

[0019] According to the present invention, collecting specific data on fish farming has the effect of aiding various future research on fish.

[0020] The effects of the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0022] Figure 1 illustrates the structure of a feed quantity prediction model according to an embodiment of the present invention. Figure 2 illustrates the structure of a feed amount and growth model according to an embodiment of the present invention. FIG. 3 illustrates an overall flowchart of fish productivity management according to an embodiment of the present invention. Figure 4 illustrates a method for managing fish productivity in land-based aquaculture farms through data prediction by growth stage. FIG. 5 is a block diagram showing a computer system for implementing a method according to an embodiment of the present invention. Specific details for implementing the invention

[0023] The aforementioned objectives of the present invention, as well as other objectives, advantages, and features, and the methods for achieving them, will become clear from the embodiments described in detail below together with the accompanying drawings.

[0024] However, the present invention is not limited to the embodiments disclosed below but can be implemented in various different forms, and the following embodiments are provided merely to easily inform those skilled in the art of the purpose, structure, and effects of the invention, and the scope of the rights of the present invention is defined by the description in the claims.

[0025] Meanwhile, the terms used in this specification are for describing the embodiments and are not intended to limit the invention. In this specification, the singular form includes the plural form unless specifically stated otherwise in the text. As used in this specification, "comprises" and / or "comprising" do not exclude the presence or addition of one or more other components, steps, actions, and / or elements to the mentioned components, steps, actions, and / or elements.

[0026] According to an embodiment of the present invention, feed input data is automatically predicted through sensor data installed in a tank, and an automated fish growth model is presented by periodically correcting it, thereby performing stable modeling by ensuring the stability of the input feed amount data.

[0027] According to an embodiment of the present invention, a fish growth model divided by period is presented to address data loss caused by the sorting process, which is applied to an aquaculture farm in the field that periodically places individuals of the same size into a single tank. According to an embodiment of the present invention, the dispersion of fish body weight within the tank according to fish growth is predicted, and based on this, the timing of sorting is proposed.

[0028] According to an embodiment of the present invention, a sensor is installed in a land-based aquaculture tank, and the trend of changes in data generated when feeding is predicted to estimate the amount of feed and thereby construct a growth model. According to an embodiment of the present invention, by analyzing the sorting process of an actual fish farm, it is possible to construct a growth model with improved accuracy in a segmented form according to fish growth.

[0029] Figure 1 illustrates the structure of a feed quantity prediction model according to an embodiment of the present invention.

[0030] The feed amount prediction model predicts the amount of feed to be input based on data such as environmental data installed in the tank, the width and height of the tank, the amount of water input sensed through flow sensors, and the biomass of the fish in the tank.

[0031] For feed quantity prediction modeling, mathematical modeling based on measurement data must be performed in advance, with the actual biomass of the fish, dissolved oxygen, water temperature, and feed quantity used as input values. The amount of dissolved oxygen consumed by the fish is measured.

[0032] It is possible to model the amount of feed supplied.

[0033] delete

[0034] delete

[0035] It is possible to estimate the amount of feed supplied according to the example above, and as another example, it is also possible to configure it in the form of a look-up table (LUT) between the amount of feed supplied and the amount of oxygen consumed.

[0036] According to an embodiment of the present invention, the error between the predicted and actual values ​​that occurs over time is reduced by performing a model update automatically by comparing the actual amount of feed input with the estimated feed amount (predicted feed amount) at regular intervals. Through this, even if the user does not manually record the feed amount, the amount of feed input can be predicted based on the trend of changes in sensing data, such as the amount of water input and the decrease in oxygen saturation.

[0037] The estimated feed supply is stored on storage devices, such as management servers, to secure data for fish production management. By simultaneously managing the actual feed supplied during operation and the feed supply estimated by the prediction model, this information can be utilized for operational management decisions. Additionally, the estimated feed supply is input into a future growth prediction model and used to forecast the future growth rate of the fish.

[0038] According to an embodiment of the present invention, in order to solve the problem of data loss due to sorting operations at the aquaculture site and the decrease in accuracy of the growth model caused by periodic tank information reset, the fish growth model is configured with a structure divided by period. According to an embodiment of the present invention, by dividing the growth of fish into segments, learning is performed, and then integrating the results, an appropriate model is applied according to the current growth level of the fish to predict the growth level, thereby improving the accuracy of the overall growth level prediction and compensating for errors caused by periodic fish sorting.

[0039] Figure 2 illustrates the structure of a feed amount and growth model according to an embodiment of the present invention.

[0040] Information such as environment data and tank biomass is input into both the feed amount prediction model and the growth prediction model, and the predicted feed amount is input into the growth prediction model to ultimately predict the growth of the fish.

[0041] The divided growth prediction model can be implemented using a time-series deep learning engine such as LSTM, and the growth prediction model is operated in parallel at the weight level. The input data for the prediction model consists of environmental data such as stocking weight, water temperature, and DO, as well as operational data such as feed quantity. The output data consists of the average growth weight and deviation values ​​after a specific time interval. The predicted average growth weight and deviation are utilized for production management to distribute fish and control shipment volumes at specific points in time.

[0042] The growth prediction model outputs predicted weight values ​​and corresponding weight variance values. The weight variance value serves as a parameter for fish size and weight, which vary due to differences in growth rates even with the same feed input, thereby enabling the prediction of the sorting time.

[0043] FIG. 3 illustrates an overall flowchart of fish productivity management according to an embodiment of the present invention.

[0044] Data regarding the aquaculture farm environment, such as environmental data from the tanks and information about the fish contained within, is entered, and handwritten data is periodically entered into the database. The handwritten data includes feed amounts and measured fish weights.

[0045] Data accumulated in the database is input into a feed amount prediction model to predict the amount of feed input and the amount to be input in the future, and data regarding the timing of fish selection and growth level is output through a growth prediction model.

[0046] FIG. 4 illustrates a method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage according to an embodiment of the present invention.

[0047] A method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage according to an embodiment of the present invention includes a step of performing a prediction on the amount of feed input (S410) and a step of performing a prediction on fish growth using the result of the prediction on the amount of feed input (S420).

[0048] In step S410, the amount of feed input is predicted using the fish body weight, initial oxygen amount, water temperature, and oxygen consumed.

[0049] As another example, in step S410, a prediction of the amount of feed input can be performed using a lookup table for the amount of oxygen consumed and the amount of feed supplied.

[0050] In step S410, the feed quantity prediction model is updated by comparing the predicted result of the input feed quantity with the actual input feed quantity.

[0051] In step S420, growth prediction models divided by weight are operated in parallel to perform predictions on fish growth.

[0052] In step S420, a prediction for the sorting time is performed using the predicted weight value and weight dispersion value.

[0053] The method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage according to an embodiment of the present invention further includes the step (S430) of transmitting optimal growth conditions according to the fish growth prediction result.

[0054] In step S430, control elements for the environmental data within the tank are calculated, and control commands are transmitted accordingly. For example, it is possible to transmit control commands to adjust the amount of feed supplied, the amount of dissolved oxygen, or the water temperature to satisfy environmental conditions for each growth stage. Optimal growth conditions can be set differently depending on various production conditions. For instance, optimal growth conditions are set for each period by comprehensively considering factors such as the shipping time, estimated shipping volume, information on expected price fluctuations by period, information on changes in price trends, and information on expected price trends, and control commands are transmitted accordingly.

[0055] FIG. 5 is a block diagram showing a computer system for implementing a method according to an embodiment of the present invention.

[0056] Referring to FIG. 5, a computer system (1000) may include at least one of a processor (1010), memory (1030), an input interface device (1050), an output interface device (1060), and a storage device (1040) that communicate via a bus (1070). The computer system (1000) may also include a communication device (1020) coupled to a network. The processor (1010) may be a central processing unit (CPU) or a semiconductor device that executes instructions stored in memory (1030) or a storage device (1040). Memory (1030) and storage device (1040) may include various forms of volatile or non-volatile storage media. For example, memory may include read-only memory (ROM) and random access memory (RAM). In the embodiments of this description, memory may be located inside or outside the processor, and memory may be connected to the processor through various known means. Memory is a volatile or non-volatile storage medium of various forms, and for example, memory may include read-only memory (ROM) or random access memory (RAM).

[0057] Accordingly, embodiments of the present invention may be implemented as a method implemented on a computer or as a non-transient computer-readable medium storing computer-executable instructions. In one embodiment, when executed by a processor, the computer-readable instructions may perform a method according to at least one aspect of the present description.

[0058] The communication device (1020) can transmit or receive wired or wireless signals.

[0059] In addition, the method according to an embodiment of the present invention may be implemented in the form of program instructions that can be executed through various computer means and may be recorded on a computer-readable medium.

[0060] The above computer-readable medium may include program instructions, data files, data structures, etc., either individually or in combination. The program instructions recorded on the computer-readable medium may be specially designed and configured for embodiments of the present invention, or they may be known and available to a person skilled in the art of computer software. The computer-readable recording medium may include a hardware device configured to store and execute program instructions. For example, the computer-readable recording medium may be magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; ROM; RAM; flash memory, etc. The program instructions may include not only machine code, such as that generated by a compiler, but also high-level language code that can be executed by a computer through an interpreter, etc.

[0061] Although embodiments of the present invention have been described in detail above, the scope of the present invention is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present invention as defined in the following claims also fall within the scope of the present invention.

Claims

Claim 1 A method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage performed by a fish productivity management device for a land-based aquaculture farm, comprising: (a) a step of predicting the amount of feed input into a tank using tank sensing data; (b) a step of predicting fish growth using the result of the prediction of the amount of feed input; and (c) a step of calculating control elements for environmental data within the tank using the result of the prediction of fish growth and transmitting control commands to control the amount of feed supplied, dissolved oxygen, and water temperature, wherein step (b) performs the prediction of fish growth by operating growth prediction models divided into weight units in parallel, predicting the growth gap that occurs during fish sorting, and outputting weight values ​​and weight dispersion values ​​using the growth prediction models to predict the sorting time; and step (c) transmits the control commands according to growth conditions set by time, considering the time of shipment, estimated shipment quantity, information on price fluctuations by time, information on price trend changes, and information on price trend prediction. Claim 2 delete Claim 3 A method for managing fish productivity in a land-based aquaculture farm through data prediction by growth stage, wherein step (a) involves performing a prediction of the amount of feed input using a lookup table for the amount of oxygen consumed and the amount of feed supplied. Claim 4 delete Claim 5 delete Claim 6 A land-based aquaculture fish productivity management device comprising: an input unit receiving input information including tank information and environmental information; a memory storing a program that performs a prediction of the amount of feed input using the input information; and a processor that executes the program, wherein the processor divides a growth prediction model by sorting time considering the prediction result of the amount of feed input and outputs the prediction result of the fish growth rate and sorting time, and the processor performs a prediction of fish growth by operating a growth prediction model divided by weight unit in parallel, predicts the growth gap that occurs during fish sorting, and performs a prediction of the sorting time by outputting weight values ​​and weight dispersion values ​​using the growth prediction model, and the processor calculates control elements for environmental data within the tank using the prediction result and transmits control commands to adjust the amount of feed supplied, dissolved oxygen amount, and water temperature, wherein the control commands are transmitted according to growth conditions set by time considering the shipment time, estimated shipment amount, information on price fluctuations by time, information on price trend change, and information on price trend prediction.

Images