Data-based flow-type smart aquaculture farm automated water quality management system and operation method thereof

Abstract

A flow-through smart aquaculture automated water quality management system comprises: a plurality of aquaculture tanks containing water for cultivating aquatic products; a plurality of pipes connecting the plurality of aquaculture tanks; a plurality of valves installed in the plurality of pipes to regulate or block the flow of water within the plurality of pipes; a plurality of pumps installed in the plurality of pipes to provide pressure for the flow of water within the plurality of pipes; a tank sensor installed in the plurality of aquaculture tanks to acquire tank environment data regarding the environment of the plurality of aquaculture tanks; an image sensor installed in the plurality of aquaculture tanks to acquire aquatic product data regarding the aquatic products; a communication device to acquire external environment data regarding seawater flowing into the plurality of aquaculture tanks; an analysis server that analyzes the tank environment data, the aquatic product data, and the external environment data to output an analysis result and generates a control signal according to the analysis result; a monitoring device that provides the analysis result from the analysis server as visualization data; a valve gateway that controls the plurality of valves based on the control signal received from the analysis server; a pump gateway that controls the plurality of pumps based on the control signal received from the analysis server; and a water quality treatment system that processes the water contained in the plurality of aquaculture tanks based on the control signal received from the analysis server. Includes a device.

Description

Technology Field

[0001] The technical field to which the present invention belongs relates to an automated water quality management system for a flow-through smart fish farm, and more specifically, to a system that predicts changes in water quality of a flow-through smart fish farm by utilizing tank environment data and external environment data, and saves energy through optimal pump control. Background Technology

[0002] The content described in this section merely provides background information regarding the present embodiment and does not constitute prior art.

[0003] With the recent surge in global seafood consumption, the sustainable aquaculture industry is becoming increasingly important. As traditional fishing methods alone are insufficient to meet this growing demand, the development of the aquaculture industry for stable and sustainable fish production is essential. In particular, due to recent environmental issues and the depletion of marine resources, the aquaculture industry is now faced with the challenge of transforming into environmentally friendly and energy-efficient practices.

[0004] The introduction of smart aquaculture systems marked a significant turning point in resolving the various problems that the aquaculture industry had been facing due to traditional operational methods. Since existing aquaculture management methods were largely passive, relying on the intuition or experience of the operators, it was difficult to respond effectively to changes in the external environment, and operational efficiency was also limited. To overcome this, continuous efforts were made to combine existing traditional aquaculture techniques with the latest cutting-edge technologies, and as a result, the concept of smart aquaculture has established itself as a core direction for the aquaculture industry in recent years.

[0005] Smart aquaculture actively utilizes the latest cutting-edge technologies, including artificial intelligence (AI), the Internet of Things (IoT), and facility automation, as well as the comprehensive analysis of data collected throughout the entire growth process. This approach goes beyond mere technical improvements and places a primary goal on enhancing both the sustainability and economic viability of the aquaculture industry. In particular, the core strength of smart aquaculture systems lies in providing practical and concrete support to operators by integrating existing methods with the latest technologies to increase operational efficiency, minimize downtime, and systematically optimize maintenance practices.

[0006] The environment of aquaculture farms is constantly changing due to complex and diverse external factors, such as seasonal shifts, fluctuations in weather conditions, and seawater temperature. In contrast, however, the internal environment of the farm must be managed in a very strict and stable state to ensure the health and survival of the fish. This is because even the slightest change in the internal environment poses a risk of severe stress or mortality. In other words, environmental management in aquaculture farms needs to be advanced to enable a rapid and accurate response to sudden changes in the external environment.

[0007] Aquaculture farms are broadly classified into flow-through and circulation-through farms.

[0008] Flow-through fish farms are designed to continuously draw in and discharge external seawater. A stable water circulation within the tanks is essential to ensure a healthy growth environment for farmed fish. In this process, pumps are critical equipment for maintaining a constant flow of inflow and outflow water. As pump operation time increases, power consumption also rises. In particular, since pump loads vary with changes in seawater tidal levels, simply relying on fixed-cycle, time-based scheduling directly leads to excessive energy consumption.

[0009] Feeding, a key factor in fish growth, is linked to water quality issues. Feeding accelerates the accumulation of organic matter (such as residual feed and sediment) in aquaculture tanks, which leads to a decrease in dissolved oxygen (DO) concentration and an increase in pollutants like ammonia and nitrates. Consequently, water quality deterioration has a direct adverse effect on fish health and can increase mortality rates; therefore, the timing of feed supply and water quality maintenance must be closely coordinated. It is necessary to implement control measures that appropriately combine oxygen supply and water circulation within a certain period after feeding.

[0010] Existing systems utilize a simple feedback control structure that primarily activates pumps or oxygen suppliers when a certain threshold is exceeded, based on real-time data from water quality sensors. This approach is limited to responding after an event occurs, making it difficult to proactively predict and respond to changes in water quality caused by weather conditions or feed timing.

[0011] With regard to flow-through fish farms, Korean Registered Patent No. 10-2743323 describes a flow-through fish farm equipped with a filtration function, and Korean Registered Patent No. 10-2505704 describes an aquaculture system that enables the operation of an aging flow-through fish farm into a flow-through and recirculating filtration type fish farm. Korean Registered Patents No. 10-2743323 and No. 10-2505704 do not consider a configuration for automatically managing the water quality of a flow-through fish farm based on data. Prior art literature

[0012] Korean Registered Patent No. 10-2743323 (December 16, 2024) Korean Registered Patent No. 10-2505704 (March 3, 2023) The problem to be solved

[0013] The embodiments of the present invention are an energy data-based integrated control system designed to simultaneously maintain a balance between optimizing energy consumption and ensuring water quality stability. The primary purpose is to overcome the limitations of existing flow-through aquaculture farms and improve efficiency and stability by integrating water quality monitoring, pump operation optimization, water quality time-series prediction models, and remote control technology.

[0014] Other unspecified objects of the present invention may be further considered to the extent that they can be easily inferred from the following detailed description and effects. means of solving the problem

[0015] According to one aspect of the present embodiment, a water quality management system for an automated smart aquaculture farm with a flow-through system comprises: a plurality of aquaculture tanks containing water for cultivating aquatic products; a plurality of pipes connecting the plurality of aquaculture tanks; a plurality of valves installed in the plurality of pipes to regulate or block the flow of water within the plurality of pipes; a plurality of pumps installed in the plurality of pipes to provide pressure for water to flow within the plurality of pipes; a tank sensor installed in the plurality of aquaculture tanks to acquire tank environment data regarding the environment of the plurality of aquaculture tanks; an image sensor installed in the plurality of aquaculture tanks to acquire aquatic product data regarding the aquatic products; a communication device to acquire external environment data regarding seawater flowing into the plurality of aquaculture tanks; an analysis server that analyzes the tank environment data, the aquatic product data, and the external environment data to output an analysis result and generates a control signal according to the analysis result; a monitoring device that provides the analysis result from the analysis server as visualization data; and a valve gateway that controls the plurality of valves based on the control signal received from the analysis server. The system includes a pump gateway that controls the plurality of pumps based on the control signal received from the analysis server, and a water quality treatment device that treats the water contained in the plurality of aquaculture tanks based on the control signal received from the analysis server, comprising a flow-through smart aquaculture farm automated water quality management system.

[0016] The analysis server measures the flow rate of the plurality of pumps, the rotational speed of the plurality of pumps, and the power consumption of the plurality of pumps through the pump gateway, determines the operation schedule of the plurality of pumps to satisfy minimum power consumption and flow rate conditions according to the relationship between the flow rate of the plurality of pumps, the rotational speed of the plurality of pumps, and the power consumption of the plurality of pumps, and can include information regarding the operation schedule of the plurality of pumps in the control signal and transmit it to the pump gateway.

[0017] The analysis server predicts the water quality results of the plurality of aquaculture tanks using an LSTM (Long Short-Term Memory) hybrid model and a regression analysis model based on the external environment data and the tank environment data, and in the LSTM hybrid model, in an LSTM encoder-LSTM decoder combined structure based on an encoder-decoder structure, when the LSTM decoder predicts the water quality results, attention using the external environment data is applied to the LSTM encoder, and the analysis server can include the predicted water quality results in the control signal and transmit them to the water quality treatment device.

[0018] The analysis server communicates with the pump gateway and the water quality treatment device using MQTT (Message Queuing Telemetry Transport) and MODBUS protocols, the pump gateway controls the plurality of pumps according to a control signal containing information regarding the operation schedule, and the water quality treatment device can control the oxygen supply according to a control signal containing the predicted water quality result.

[0019] The analysis server analyzes the tank environment data, the aquatic product data, and the external environment data to select one of a basic mode, a filtration mode, or a circulation mode, includes the selected mode in the control signal, and transmits it to the valve gateway, and the valve gateway can change the connection path of the plurality of pipes by opening and closing the plurality of valves according to the control signal including the selected mode.

[0020] The analysis server selects the first aquaculture tank among the plurality of aquaculture tanks as the filtration tank in the filtration mode, and the valve gateway can change the connection path of the plurality of pipes by opening and closing the plurality of valves so that the incoming water flows through the first aquaculture tank to another aquaculture tank and is discharged to the outside.

[0021] The analysis server can change the connection path of the plurality of pipes by opening and closing the plurality of valves so that water flows through the first aquaculture tank to the second aquaculture tank, water flows from the second aquaculture tank to another aquaculture tank, and water flows back into the first aquaculture tank from the other aquaculture tank in the circulation mode, and the valve gateway can change the connection path of the plurality of pipes.

[0022] According to another aspect of the present embodiment, a method of operation of a flow-through smart aquaculture automated water quality management system may include: a step of acquiring tank environment data regarding the environment of a plurality of aquaculture tanks through tank sensors installed in a plurality of aquaculture tanks; a step of acquiring aquatic product data regarding aquatic products through image sensors installed in the plurality of aquaculture tanks; a step of acquiring external environment data regarding seawater flowing into the plurality of aquaculture tanks through a communication device; a step of analyzing the tank environment data, the aquatic product data, and the external environment data through an analysis server to output an analysis result and generate a control signal according to the analysis result; a step of providing the analysis result as visualization data through a monitoring device; a step of controlling a plurality of valves installed in the plurality of pipes based on the control signal received from the analysis server through a valve gateway; a step of controlling a plurality of pumps installed in the plurality of pipes based on the control signal received from the analysis server through a pump gateway; and a step of treating the water contained in the plurality of aquaculture tanks based on the control signal received from the analysis server through a water quality treatment device. Effects of the invention

[0023] As described above, according to the embodiments of the present invention, using tide level prediction data and water quality data, integrated water quality prediction and operation control based on multiple regression analysis and LSTM hybrid network is provided, and there is an effect of being able to quantitatively analyze complex interactions between water quality, feed, and energy and implement an automation structure capable of proactive response.

[0024] Even if an effect is not explicitly mentioned herein, the effects described in the following specification and other effects that are expected by the technical features of the present invention and can be easily inferred are treated as described in the specification of the present invention. Brief explanation of the drawing

[0025] FIG. 1 is a diagram illustrating the configuration of an automated water quality management system for a smart fish farm according to one embodiment of the present invention. FIGS. 2 and FIGS. 3 are drawings illustrating each operation of a flow-through smart fish farm automated water quality management system according to one embodiment of the present invention. FIGS. 4 to 7 are drawings illustrating a pipe connection structure according to a mode of a flow-through smart fish farm automated water quality management system according to an embodiment of the present invention. FIG. 8 is a flowchart illustrating the operation method of an automated water quality management system for a smart fish farm according to one embodiment of the present invention. Specific details for implementing the invention

[0026] Hereinafter, in describing the present invention, detailed descriptions of related known functions are omitted if they are deemed obvious to a person skilled in the art and could unnecessarily obscure the essence of the invention, and some embodiments of the present invention are described in detail through exemplary drawings.

[0027] FIG. 1 is a diagram illustrating the configuration of an automated water quality management system for a smart fish farm according to one embodiment of the present invention.

[0028] The automated water quality management system (100) for a smart fish farm with a flowing water system includes an automation system and an operation algorithm to simultaneously ensure operational efficiency and water quality stability of the fish farm. Basically, by analyzing the potential meaning of data through the correlation of complex data, it predicts changes in water quality and provides an algorithm that saves energy through optimal pump control accordingly.

[0029] The tidal level data processed by the flow-through smart fish farm automated water quality management system (100) is collected from a public API (Application Programming Interface), and when the tidal level rises, the point at which a high flow rate can be secured with relatively low energy is analyzed, and the operating rate of the pump is adjusted accordingly. In this process, the power consumption and water tank flow rate according to the change in pump rotation speed are quantified to establish an operation schedule that achieves the maximum water circulation effect with the minimum power consumption.

[0030] The water quality prediction model applied to the flow-through smart aquaculture automated water quality management system (100) derives a correlation by comprehensively analyzing past sensor data, water quality data, and external climate factors, and can predict signs of water quality abnormalities such as a rise in water temperature, a decrease in dissolved oxygen (DO), changes in pH, and an increase in turbidity within a specific time period after feeding. Through this, when feed is scheduled to be supplied at a specific time, the system takes precautionary measures to prevent water quality deterioration by adjusting the amount of oxygen supplied or the amount of inflow water in advance.

[0031] The flow-through smart fish farm automated water quality management system (100) operates by organically linking water circulation through pump control and water quality prediction to reduce energy consumption and maintain stable water quality. Based on the collected data, the system optimizes water quality management of the flow-through fish farm, enabling stable operation of the fish farm.

[0032] Referring to FIG. 1, the automated water quality management system (10) for a flowing smart fish farm includes a plurality of tank sensors (111), an image sensor (112), a communication device (113), a plurality of aquaculture tanks (120), a plurality of valves (130), a valve gateway (135), a plurality of pumps (140), a pump gateway (145), an analysis server (150), a monitoring device (155), and a water quality treatment device (160). The plurality of tank sensors (111), the image sensor (112), the communication device (113), the plurality of aquaculture tanks (120), the plurality of valves (130), the valve gateway (135), the plurality of pumps (140), the pump gateway (145), the analysis server (150), the monitoring device (155), and the water quality treatment device (160) are interconnected to transmit and receive data or control signals related to the flowing smart fish farm.

[0033] Multiple tank sensors (111) are installed in multiple aquaculture tanks to acquire tank environment data regarding the environment of multiple aquaculture tanks. For example, tank water temperature, DO, pH, etc. can be acquired.

[0034] The image sensor (112) is installed in a plurality of aquaculture tanks to acquire aquatic product data regarding aquatic products. For example, it can be implemented as an underwater camera, etc.

[0035] The communication device (113) obtains external environmental data regarding seawater flowing into multiple aquaculture tanks. For example, it can obtain seawater temperature, tide level, etc.

[0036] Multiple aquaculture tanks (120) are tanks containing water for cultivating aquatic products. Multiple aquaculture tanks (120) are connected through multiple pipes (125).

[0037] Multiple valves (130) are installed in multiple pipes to control or block the flow of water within the multiple pipes.

[0038] The valve gateway (135) controls the opening and closing of multiple valves based on control signals received from the analysis server.

[0039] Multiple pumps (140) are installed in multiple pipes to provide pressure so that water flows in multiple pipes.

[0040] The pump gateway (145) controls the on / off and speed of multiple pumps based on control signals received from the analysis server.

[0041] The analysis server (150) analyzes the tank environment data, fishery product data, and external environment data, outputs the analysis results, and generates a control signal according to the analysis results.

[0042] The monitoring device (155) provides the analysis results received from the analysis server as visualization data.

[0043] The water quality treatment device (160) processes the water contained in multiple aquaculture tanks based on a control signal received from an analysis server.

[0044] The analysis server (150) measures the flow rate of multiple pumps, the rotational speed of multiple pumps, and the power consumption of multiple pumps through the pump gateway. The analysis server (150) determines the operation schedule of multiple pumps to satisfy minimum power consumption and flow rate conditions based on the relationship between the flow rate of multiple pumps, the rotational speed of multiple pumps, and the power consumption of multiple pumps. The analysis server (150) includes information regarding the operation schedule of multiple pumps in a control signal and transmits it to the pump gateway.

[0045] The analysis server (150) predicts the water quality results of multiple aquaculture tanks using an LSTM (Long Short-Term Memory) hybrid model and a regression analysis model based on external environment data and tank environment data. In the LSTM hybrid model, in an LSTM encoder-LSTM decoder combined structure based on an encoder-decoder structure, the LSTM decoder applies attention using external environment data to the LSTM encoder when predicting the water quality results. The analysis server (150) includes the predicted water quality results in a control signal and transmits it to a water quality treatment device.

[0046] The analysis server (150) communicates with the pump gateway (145) and the water quality treatment device (160) using MQTT (Message Queuing Telemetry Transport) and MODBUS protocols. The pump gateway (145) controls the on / off and speed of a plurality of pumps according to a control signal containing information regarding the operation schedule.

[0047] The water quality treatment device (160) controls the oxygen supply according to a control signal containing predicted water quality results. The water quality treatment device (160) can control the oxygen supply to a plurality of aquaculture tanks, etc.

[0048] FIGS. 2 and FIGS. 3 are drawings illustrating each operation of a flow-through smart fish farm automated water quality management system according to one embodiment of the present invention.

[0049] Referring to Fig. 2, the automated water quality management system for a smart fish farm collects environmental data. It collects data regarding water temperature, DO, salinity, pH, turbidity, etc. through sensors installed in the fish farm. It also collects data such as tide level and weather information using an external API.

[0050] The flow-through smart aquaculture automated water quality management system integrates data and detects outliers. Data is stored on an analysis server, which can perform normalization through preprocessing to remove missing values ​​and outlier data.

[0051] The automated water quality management system for smart aquaculture farms predicts water quality. Through LSTM hybrid models and regression analysis, it analyzes changes in water quality after feeding, predicts water quality time series data, and can predict changes such as DO reduction, pH changes, and turbidity.

[0052] The flow-through smart aquaculture automated water quality management system conducts energy simulations. The analysis server controls fluctuations in pump drive rates and generates an optimal power operation schedule.

[0053] The flow-through smart aquaculture automated water quality management system generates control commands. The analysis server determines water quality thresholds and generates commands for prediction-based pump / oxygen supply control.

[0054] The flow-through smart aquaculture automated water quality management system remotely controls each piece of equipment. The analysis server can automatically control pumps and oxygenators using the MQTT / Modbus protocol.

[0055] The flow-through smart aquaculture automated water quality management system provides feedback through a dashboard. The monitoring device visualizes real-time status and presents dashboard information to the user. It provides an interface that allows for alarms and manual user intervention depending on the situation.

[0056] Referring to Fig. 3, the automated water quality management system for a flow-through smart aquaculture farm applies energy-optimized pump control technology. Due to the structural characteristics of flow-through aquaculture farms, which maintain water quality by continuously inflowing and outflowing seawater, the constant operation of pumps is unavoidable. Consequently, energy consumption accounts for a significant portion of the total operating costs of the aquaculture farm, and in particular, the load on the pumps fluctuates frequently depending on external environmental factors such as tidal changes, water temperature, and flow velocity. To address these issues, this study provides a pump operation schedule optimization algorithm based on tidal prediction data.

[0057] The automated water quality management system for smart aquaculture farms receives real-time tide level prediction data from a public API, and since the pump load decreases when the tide level is high, power consumption can be reduced under the same flow rate conditions by concentrating pump operations around this time period.

[0058] The analysis server of the automated water quality management system for flow-through smart aquaculture farms provides an operation simulator that numerically models the physical characteristics of pumps (relationships between flow rate, rotational speed, and power consumption) and quantitatively calculates energy consumption based on pump rotational speed. Subsequently, it derives an operation schedule that simultaneously satisfies the conditions for minimum energy consumption and sufficient flow rate under given tidal conditions. This schedule is transmitted to the pump units via MQTT or MODBUS protocols for automatic real-time execution, allowing operators to monitor the operation status and estimated power savings through a dashboard.

[0059] The flow-through smart fish farm automated water quality management system provides a water quality prediction algorithm.

[0060] Water quality within aquaculture farms fluctuates rapidly under the influence of various factors, such as feed supply, changes in external seawater temperature, wind direction, and rainfall, which can lead to fish mortality or stunted growth. In particular, organic matter concentrations temporarily increase after feeding, resulting in phenomena such as oxygen consumption and a decrease in pH. Since it is difficult to respond to these complex water quality changes by simply exceeding standard values, the automated water quality management system for flow-through smart aquaculture farms provides a structure capable of proactive response through prediction algorithms based on water quality time-series data.

[0061] The prediction algorithm applied to the automated water quality management system for flow-through smart aquaculture farms is based on data such as water temperature, dissolved oxygen (DO), salinity, pH, and turbidity, including historical and real-time sensor data, and also utilizes external meteorological data such as tide level, seawater temperature, air temperature, and wind speed as input values. The algorithm structure is configured to linearly analyze key causal relationships through multiple linear regression, and then learn long-term patterns and non-linear variability within the time series using a hybrid network combining LSTM (Long Short-Term Memory), Seq2Seq, and Attention.

[0062] Through this, if it is determined that predicted water quality indicators will reach a pre-set threshold, the system outputs control commands, such as increasing the oxygen supply or adjusting the inflow volume, if necessary. At this time, combined with energy-optimized pump control technology, the system manages water quality by controlling the flow rate to the optimal level for water quality management based on the tidal level. This prediction-based control enables proactive water quality management rather than simple reactive responses, thereby minimizing fish damage caused by water quality abnormalities and supporting the maintenance of a stable aquaculture environment.

[0063] The flow-through smart aquaculture automated water quality management system collects spatiotemporal environmental data based on multiple sources. It collects tank temperature (T), dissolved oxygen (DO), salinity (S), pH, turbidity, flow velocity, and flow rate through sensors. It also collects tide level (H), red tide, and sea water temperature via APIs.

[0064] The flow-through smart aquaculture automated water quality management system performs adaptive data preprocessing and intelligent clustering. The analysis server can filter out noise and outliers using Gaussian Mixture Models (GMM) and Mahalanobis Distance. The analysis server can perform automatic clustering by tank / environment type. For example, DBSCAN or a modified k-means algorithm can be applied. Data MinMaxScalar normalization can be performed.

[0065] The automated water quality management system for smart aquaculture farms can apply a high-resolution water quality prediction algorithm with a multi-model fusion structure. The analysis server adopts a hybrid structure, applies LSTM for short-term prediction and Seq2Seq and Attention for medium-term prediction (6–24 hours), performs multi-model switching based on situational comparison criteria through regression ensemble, and can utilize external factors as Attention Queries.

[0066] Mathematical Equation 1 is an example of a prediction regarding LSTM.

[0067] [Mathematical Formula 1]

[0068]

[0069] Mathematical Equation 2 illustrates Seq2Seq and Attention-based prediction.

[0070] [Mathematical Formula 2]

[0071]

[0072] Mathematical Equation 3 illustrates an adjusted multiple regression considering environment-feeding information variables.

[0073] [Mathematical Formula 3]

[0074]

[0075] T is the tank temperature, Q is the flow rate (return volume), F is the feed amount (feed supply amount per unit of time in the tank), H is the tide level, pH is the hydrogen ion concentration index, SST is the open sea temperature, and ε is the error term.

[0076] The automated water quality management system for flow-through smart aquaculture farms can interpret prediction results. The analysis server analyzes risk based on DO / pH boundary values ​​and can predict water temperature and DO responses to changes in pump flow rate.

[0077] The flow-through smart aquaculture automated water quality management system can generate automatic control commands. It can generate and apply multi-condition-based control commands combined with optimal pump operation requirements.

[0078] Mathematical Equation 4 relates to determining the pump driving rate and combines the tide level, predicted flow rate, and DO prediction results.

[0079] [Mathematical Formula 4]

[0080]

[0081] is an estimated value of the return amount required to maintain water quality, and is combined with the results of the pump control optimization algorithm based on the predicted value. RPM for pump control count Convert to, and convert RPM to frequency.

[0082] The analysis server calculates energy consumption and generates an optimal operation schedule that reflects the current tide level (high / low).

[0083] The flow-through smart fish farm automated water quality management system provides an interface for real-time monitoring of execution results.

[0084] The automated water quality management system for flow-through smart aquaculture farms implements an IoT sensor network (temperature, dissolved oxygen, salinity, pH, and turbidity sensors) through a monitoring visualization panel for data display and integrates external API data. It provides a data integration and visualization dashboard and links external API data (tide level, weather data, etc.) with the Korea Hydrographic and Oceanographic Agency server and others. It collects real-time environmental data for each tank, transmits it to a server, displays it on a dashboard, triggers alarms and records abnormal values, and subsequently processes the data comprehensively on an analysis server equipped with a simulator and control module.

[0085] The flow-through smart aquaculture automated water quality management system can establish a domain for data facility control. It uses MQTT, MODBUS TCP, and MODBUS RTU (485) protocols, uses variable frequency drive inverter-type pumps, applies oxygen supply pump and feed dispenser interfaces, and applies a sliding control interface (manual adjustment of pump rotation speed is possible).

[0086] The flow-through smart aquaculture automated water quality management system receives pump operation schedules and water quality predictions provided by a simulator, automatically adjusts pump driving rates and oxygen supply levels, and transmits data and control signals in the sequence of server-gateway-equipment (valves, pumps). The control status is displayed on a dashboard in real time, and immediate alarms and notifications are provided in the event of failure.

[0087] FIGS. 4 to 7 are drawings illustrating a pipe connection structure according to a mode of a flow-through smart fish farm automated water quality management system according to an embodiment of the present invention.

[0088] Referring to FIG. 4, a plurality of aquaculture tanks may include a first aquaculture tank (211), a second aquaculture tank (212), a third aquaculture tank (213), and a fourth aquaculture tank (214). The first aquaculture tank (211), the second aquaculture tank (212), the third aquaculture tank (213), and the fourth aquaculture tank (214) are connected through pipes.

[0089] A plurality of valves connected to the pipe may include a first-1 valve (311), a first-2 valve (312), a first-3 valve (313), a first-4 valve (314), a first-5 valve (315), a first-6 valve (316), and a first-7 valve (317). The first-1 valve (311) is installed at the inlet.

[0090] A plurality of valves connected to the pipe may include a second-1 valve (321), a second-2 valve (322), a second-3 valve (323), a second-4 valve (324), and a second-5 valve (325). The second-2 valve (322) is installed on one side of the first aquaculture tank (211). The second-3 valve (323) is installed on one side of the second aquaculture tank (212). The second-5 valve (325) is installed on one side of the third aquaculture tank (213).

[0091] A plurality of valves connected to the pipe may include a third-1 valve (331), a third-2 valve (332), a fourth-1 valve (341), a fourth-2 valve (342), a fourth-3 valve (343), and a fourth-4 valve (344). The third-1 valve (331) and the fourth-1 valve (341) are installed on the other side of the first aquaculture tank (211). The third-2 valve (332) and the fourth-3 valve (343) are installed on the other side of the second aquaculture tank (212).

[0092] A plurality of valves connected to the pipe may include a 5-1 valve (351), a 5-2 valve (352), a 5-3 valve (353), and a 5-4 valve (354). The 5-4 valve (354) is installed at the outlet.

[0093] A plurality of pumps connected to the pipe may include a first pump (411), a second pump (412), a third pump (413), a fourth pump (414), and a fifth pump (415). The first pump (411) is connected to the first-1 valve. The second pump (412) is connected between the second-1 valve (321) and the fifth-1 valve (351). The third pump (413) is connected to the third-1 valve (331) and is connected between the first-3 valve (313) and the first-4 valve (314). The fourth pump (414) is connected to the third-2 valve (332) and is connected between the first-6 valve (316) and the first-7 valve (317). The fifth pump (415) is connected to the fifth-4 valve (354).

[0094] The analysis server can analyze tank environment data, aquatic product data, and external environment data to select one of the following modes: basic mode, filtration mode, or circulation mode.

[0095] The analysis server includes the selected mode in a control signal and transmits it to the valve gateway, and the valve gateway can change the connection path of multiple pipes by opening and closing multiple valves according to the control signal including the selected mode.

[0096] Referring to FIG. 5, the analysis server can form a pipe connection path so that seawater flows in during the basic mode for water flow and then passes through a plurality of tanks (first aquaculture tank, second aquaculture tank, third aquaculture tank, fourth aquaculture tank) before being discharged.

[0097] The analysis server transmits a control signal to the valve gateway to open the first-1 valve (311), open the first-2 valve (312), open the first-3 valve (313), open the first-4 valve (314), open the first-5 valve (315), open the first-6 valve (316), and open the first-7 valve (317) in the pipes connected to the first aquaculture tank (211), the second aquaculture tank (212), the third aquaculture tank (213), and the fourth aquaculture tank (214).

[0098] The analysis server transmits a control signal to the valve gateway to close the second-1 valve (321), open the second-2 valve (322), open the second-3 valve (323), close the second-4 valve (324), and open the second-5 valve (325).

[0099] The analysis server transmits a control signal to the valve gateway to close the third-1 valve (331) and the third-2 valve (332).

[0100] The analysis server transmits a control signal to the valve gateway to open the 4-1 valve (341), open the 4-2 valve (342), open the 4-3 valve (343), and open the 4-4 valve (344).

[0101] The analysis server transmits a control signal to the valve gateway to close the 5-1 valve (351), open the 5-2 valve (352), open the 5-3 valve (353), and open the 5-4 valve (354).

[0102] The analysis server transmits a control signal to the pump gateway to drive the first pump (411), stop the second pump (412), stop the third pump (413), stop the fourth pump (414), and drive the fifth pump (415). A flow-through structure is formed in which water passes through the pipes and flows in and out externally.

[0103] Referring to FIG. 6, in filtration mode, the analysis server selects the first aquaculture tank as the filtration tank among a plurality of aquaculture tanks (first aquaculture tank, second aquaculture tank, third aquaculture tank, fourth aquaculture tank), and the valve gateway can change the connection path of a plurality of pipes by opening and closing a plurality of valves so that the incoming water flows through the first aquaculture tank to other aquaculture tanks (second aquaculture tank, third aquaculture tank, fourth aquaculture tank) and is discharged to the outside. That is, the first aquaculture tank (211) becomes the filtration tank, and the second aquaculture tank (212), third aquaculture tank (213), and fourth aquaculture tank (214) can be used as aquaculture tanks.

[0104] The analysis server transmits a control signal to the valve gateway to open the 1-1 valve (311), close the 1-2 valve (312), close the 1-3 valve (313), open the 1-4 valve (314), open the 1-5 valve (315), open the 1-6 valve (316), and open the 1-7 valve (317) in the pipes connected to the 1st aquaculture tank (211), the 2nd aquaculture tank (212), the 3rd aquaculture tank (213), and the 4th aquaculture tank (214).

[0105] The analysis server transmits a control signal to the valve gateway to close the second-1 valve (321), open the second-2 valve (322), open the second-3 valve (323), close the second-4 valve (324), and open the second-5 valve (325).

[0106] The analysis server transmits a control signal to the valve gateway to open the third-1 valve (331) and close the third-2 valve (332).

[0107] The analysis server transmits a control signal to the valve gateway to close the 4-1 valve (341), close the 4-2 valve (342), open the 4-3 valve (343), and open the 4-4 valve (344).

[0108] The analysis server transmits a control signal to the valve gateway to close the 5-1 valve (351), close the 5-2 valve (352), close the 5-3 valve (353), and open the 5-4 valve (354).

[0109] The analysis server transmits a control signal to the pump gateway to drive the first pump (411), stop the second pump (412), drive the third pump (413), stop the fourth pump (414), and drive the fifth pump (415). Water passes through the pipe and flows through the third-1 valve (331), forming a filterable flow-through structure that allows water to flow in and out externally.

[0110] Referring to FIG. 7, the analysis server selects the first aquaculture tank as a sedimentation tank among the plurality of aquaculture tanks (first aquaculture tank, second aquaculture tank, third aquaculture tank, fourth aquaculture tank) in circulation mode, and selects the second aquaculture tank as a filtration tank among the plurality of aquaculture tanks (first aquaculture tank, second aquaculture tank, third aquaculture tank, fourth aquaculture tank). The valve gateway can change the connection path of the plurality of pipes by opening and closing the plurality of valves so that water flows from the first aquaculture tank to the second aquaculture tank, water flows from the second aquaculture tank to other aquaculture tanks (third aquaculture tank, fourth aquaculture tank), and water flows back into the first aquaculture tank from the other aquaculture tanks (third aquaculture tank, fourth aquaculture tank). That is, the first aquaculture tank (211) can be used as a sedimentation tank, the second aquaculture tank (212) can be used as a filtration tank, and the third aquaculture tank (213) and the fourth aquaculture tank (214) can be used as aquaculture tanks.

[0111] The analysis server transmits a control signal to the valve gateway to close the first-1 valve (311), close the first-2 valve (312), close the first-3 valve (313), open the first-4 valve (314), close the first-5 valve (315), close the first-6 valve (316), and open the first-7 valve (317) in the pipes connected to the first aquaculture tank (211), the second aquaculture tank (212), the third aquaculture tank (213), and the fourth aquaculture tank (214).

[0112] The analysis server transmits a control signal to the valve gateway to open the 2-1 valve (321), open the 2-2 valve (322), open the 2-3 valve (323), open the 2-4 valve (324), and open the 2-5 valve (325).

[0113] The analysis server transmits a control signal to the valve gateway to open the third-1 valve (331) and the third-2 valve (332).

[0114] The analysis server transmits a control signal to the valve gateway to close the 4-1 valve (341), close the 4-2 valve (342), close the 4-3 valve (343), and close the 4-4 valve (344).

[0115] The analysis server transmits a control signal to the valve gateway to open the 5-1 valve (351), open the 5-2 valve (352), open the 5-3 valve (353), and close the 5-4 valve (354).

[0116] The analysis server transmits a control signal to the pump gateway to stop the first pump (411), drive the second pump (412), drive the third pump (413), drive the fourth pump (414), and stop the fifth pump (415). Water passes through the pipe and flows from the 5-1 valve (351) to the 2-1 valve (321), flows through the 3-1 valve (331), flows through the 3-2 valve (332), and a circulating structure is formed in which water is not introduced or discharged externally.

[0117] The timing for changing to filtration mode is as follows in Equation 5, and the timing for changing to circulation mode is as follows in Equation 6.

[0118] [Mathematical Formula 5]

[0119]

[0120] [Mathematical Formula 6]

[0121]

[0122] M2 is the filtration mode score, M3 is the circulation mode score, D1 is the normalized data of the tank environment, D2 is the normalized data of the aquatic product, D3 is the normalized data of the external environment, W1 is a weight considering the time at which the tank environment data was acquired, W2 is a weight considering the condition of the aquatic product based on the aquatic product data, W3 is a weight considering the season, K1 is the scaled value of the filtration mode score considering the feeding cycle in the basic mode, and K2 is the scaled value of the circulation mode score considering energy efficiency. If the filtration mode score satisfies the first range, it switches to filtration mode, and if the circulation mode score satisfies the second range, it switches to circulation mode. If the filtration mode score falls outside the first range and the circulation mode score falls outside the second range, it switches to basic mode.

[0123] It was easy to see that the accuracy of selecting a mode suitable for the environment could be improved (approximately 5-10%) by integrating tank environment data, aquatic product data, and external environment data, and switching between basic flow mode, filtration mode, and circulation mode by reflecting the time of acquisition of tank environment data, the condition of the aquatic products, the season, the feeding cycle, and energy efficiency.

[0124] FIG. 8 is a flowchart illustrating the operation method of an automated water quality management system for a smart fish farm according to one embodiment of the present invention.

[0125] The operation method of the flow-through smart aquaculture farm automated water quality management system can be performed by the flow-through smart aquaculture farm automated water quality management system, and each operation performed by the flow-through smart aquaculture farm automated water quality management system can be performed individually or the operations can be performed in combination.

[0126] In S10, a step is performed to acquire tank environment data regarding the environment of a plurality of aquaculture tanks through tank sensors installed in a plurality of aquaculture tanks.

[0127] In S20, a step of acquiring fishery product data through image sensors installed in multiple aquaculture tanks is performed.

[0128] In S30, a step is performed to acquire external environmental data regarding seawater flowing into multiple aquaculture tanks through a communication device.

[0129] In S40, the process involves analyzing tank environment data, aquatic product data, and external environment data through an analysis server to output analysis results and generating control signals based on the analysis results.

[0130] In S50, a step is performed to provide the analysis results as visualized data through a monitoring device.

[0131] In S60, a step is performed to control a plurality of valves installed in the plurality of pipes based on a control signal received from an analysis server through a valve gateway.

[0132] In S70, a step is performed to control multiple pumps installed in multiple pipes based on control signals received from the analysis server through the pump gateway.

[0133] In S80, a step is performed to treat the water contained in multiple aquaculture tanks based on control signals received from the analysis server through a water quality treatment device.

[0134] The method of operation of a flow-through smart aquaculture automated water quality management system may include the steps of measuring the flow rate of a plurality of pumps, the rotational speed of a plurality of pumps, and the power consumption of a plurality of pumps through a pump gateway; determining the operation schedule of a plurality of pumps to satisfy minimum power consumption and flow rate conditions based on the relationship between the flow rate of a plurality of pumps, the rotational speed of a plurality of pumps, and the power consumption of a plurality of pumps; and including information regarding the operation schedule of a plurality of pumps in a control signal and transmitting it to the pump gateway.

[0135] The method of operation of a flow-through smart aquaculture automated water quality management system may include the step of predicting water quality results of a plurality of aquaculture tanks using an LSTM (Long Short-Term Memory) hybrid model and a regression analysis model based on external environment data and tank environment data; the step of applying attention using external environment data to the LSTM encoder when the LSTM decoder predicts water quality results in an LSTM encoder-LSTM decoder combined structure based on an encoder-decoder structure of the LSTM hybrid model; and the step of including the predicted water quality results in a control signal and transmitting them to a water quality treatment device.

[0136] The method of operation of a flow-through smart aquaculture automated water quality management system may include the steps of communicating with a pump gateway and a water quality treatment device using MQTT (Message Queuing Telemetry Transport) and MODBUS protocols, controlling a plurality of pumps according to a control signal containing information regarding an operation schedule through the pump gateway, and controlling oxygen supply according to a control signal containing predicted water quality results through the water quality treatment device.

[0137] A method of operation for an automated water quality management system for a smart aquaculture farm may include the steps of analyzing tank environment data, aquatic product data, and external environment data to select one of a basic mode, a filtration mode, or a circulation mode; including the selected mode in a control signal and transmitting it to a valve gateway; and opening and closing a plurality of valves according to the control signal including the selected mode through the valve gateway to change the connection path of a plurality of pipes.

[0138] The method of operation of a flow-through smart aquaculture automated water quality management system may include the step of selecting a first aquaculture tank as a filtration tank among a plurality of aquaculture tanks in filtration mode, and the step of changing the connection path of a plurality of pipes by opening and closing a plurality of valves so that water introduced through a valve gateway flows through the first aquaculture tank to another aquaculture tank and is discharged to the outside.

[0139] The method of operation of a flow-through smart aquaculture automated water quality management system may include the steps of: selecting a first aquaculture tank as a sedimentation tank among a plurality of aquaculture tanks in a circulation mode; selecting a second aquaculture tank as a filtration tank among a plurality of aquaculture tanks; and changing the connection path of a plurality of pipes by opening and closing a plurality of valves so that water flows through the first aquaculture tank to the second aquaculture tank through a valve gateway, water flows from the second aquaculture tank to another aquaculture tank, and water flows back into the first aquaculture tank from the other aquaculture tank.

[0140] The pump control device (300) of the flow-through smart fish farm automated water quality management system (10) may include a processing unit, a storage unit, and a communication bus.

[0141] The processing unit can control the analysis server (150) of the flow-through smart aquaculture automated water quality management system (10) to operate. For example, the processing unit can execute one or more programs stored in the storage unit. One or more programs may include one or more computer-executable instructions, and the computer-executable instructions may be configured to cause the analysis server (150) of the flow-through smart aquaculture automated water quality management system (10) to perform operations according to exemplary embodiments when executed by the processing unit.

[0142] The storage unit is configured to store computer-executable instructions or program code, program data and / or other suitable forms of information. Computer-executable instructions or program code, program data and / or other suitable forms of information may also be provided through an input / output interface or a communication interface. A program stored in the storage unit includes a set of instructions executable by a processing unit. In one embodiment, the storage unit may be memory (volatile memory such as random access memory, non-volatile memory, or a suitable combination thereof), one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or other types of storage media that can be accessed by the analysis server (150) of the flow-through smart aquaculture automated water quality management system (10) and store desired information, or a suitable combination thereof.

[0143] The communication bus interconnects various other components of the analysis server (150) of the flow-through smart fish farm automated water quality management system (10).

[0144] The analysis server (150) of the flow-through smart aquaculture automated water quality management system (10) may also include one or more input / output interfaces and one or more communication interfaces that provide interfaces for one or more input / output devices. The input / output interfaces and communication interfaces are connected to a communication bus. Input / output devices (not shown) may be connected to other components of the analysis server (150) of the flow-through smart aquaculture automated water quality management system (10) through the input / output interfaces.

[0145] The analysis server (150) of the flow-through smart fish farm automated water quality management system (10) may be implemented within a logic circuit by hardware, firmware, software, or a combination thereof, or may be implemented using a general-purpose or specific-purpose computer. The device may be implemented using a hardwired device, a field programmable gate array (FPGA), an application-specific integrated circuit (ASIC), etc. Additionally, the device may be implemented as a system-on-chip (SoC) including one or more processors and controllers.

[0146] The analysis server (150) of the flow-through smart aquaculture automated water quality management system (10) may be mounted in the form of software, hardware, or a combination thereof on a computing device or pump control device equipped with hardware elements. The computing device or pump control device may refer to various devices including, in whole or in part, a communication device such as a communication modem for communicating with various devices or wired / wireless communication networks, a memory for storing data for executing a program, and a microprocessor for executing a program to perform calculations and commands.

[0147] Although FIG. 8 describes each process as being executed sequentially, this is merely an illustrative description, and a person skilled in the art may modify and adapt the process in various ways without departing from the essential characteristics of the embodiment of the present invention, such as changing the order described in FIG. 8, executing one or more processes in parallel, or adding other processes.

[0148] Operations according to the embodiments may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. A computer-readable medium refers to any medium that participates in providing instructions to a processor for execution. A computer-readable medium may include program instructions, data files, data structures, or a combination thereof. Examples include magnetic media, optical recording media, memory, etc. Computer programs may be distributed over networked computer systems, and computer-readable code may be stored and executed in a distributed manner. Functional programs, code, and code segments for implementing the embodiments may be easily deduced by programmers skilled in the art to which the embodiments belong.

[0149] These embodiments are intended to explain the technical concept of the present embodiment, and the scope of the technical concept of the present embodiment is not limited by these embodiments. The scope of protection of the present embodiment shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of the present embodiment. Explanation of the symbols

[0150] 100: Flow-through smart fish farm automated water quality management system 111: Water tank sensor 112: Image sensor 113: Communication device 120: Aquaculture tank 130: Valve 135: Valve Gateway 140: Pump 145: Pump Gateway 150: Analysis Server 155: Monitoring device 160: Water treatment device

Claims

Claim 1 In a flow-through smart aquaculture automated water quality management system, the system comprises: a plurality of aquaculture tanks containing water for cultivating aquatic products; a plurality of pipes connecting the plurality of aquaculture tanks; a plurality of valves installed in the plurality of pipes to regulate or block the flow of water within the plurality of pipes; a plurality of pumps installed in the plurality of pipes to provide pressure for water to flow within the plurality of pipes; a tank sensor installed in the plurality of aquaculture tanks to acquire tank environment data regarding the environment of the plurality of aquaculture tanks; an image sensor installed in the plurality of aquaculture tanks to acquire aquatic product data regarding the aquatic products; a communication device to acquire external environment data regarding seawater flowing into the plurality of aquaculture tanks; an analysis server that analyzes the tank environment data, the aquatic product data, and the external environment data to output an analysis result and generates a control signal according to the analysis result; a monitoring device that provides the analysis result from the analysis server as visualization data; a valve gateway that controls the plurality of valves based on the control signal received from the analysis server; and a pump gateway that controls the plurality of pumps based on the control signal received from the analysis server. The apparatus includes a water quality treatment device that processes water contained in the plurality of aquaculture tanks based on the control signal received from the analysis server, wherein the plurality of aquaculture tanks includes a first aquaculture tank, a second aquaculture tank, a third aquaculture tank, and a fourth aquaculture tank, and the first aquaculture tank, the second aquaculture tank, the third aquaculture tank, and the fourth aquaculture tank are connected through the plurality of pipes, and the plurality of valves include a first-1 valve, a first-2 valve, a first-3 valve, a first-4 valve, a first-5 valve, a first-6 valve, and a first-7 valve, wherein the first-1 valve is installed at an inlet, the first-2 valve is connected to the first-1 valve, and the first-3 valve is connected to the first-2 valve.The above 1-4 valve is connected to the above 1-3 valve, the above 1-5 valve is connected to the above 1-4 valve, the above 1-6 valve is connected to the above 1-5 valve, the above 1-7 valve is connected to the above 1-6 valve, and the above plurality of valves includes 2-1 valve, 2-2 valve, 2-3 valve, 2-4 valve, and 2-5 valve, the above 2-2 valve is installed on one side of the above 1 aquaculture tank, the above 2-3 valve is installed on one side of the above 2 aquaculture tank, and the above 2-5 valve is installed on one side of the above 3 aquaculture tank, and the above plurality of valves includes 3-1 valve, 3-2 valve, 4-1 valve, 4-2 valve, 4-3 valve, and 4-4 valve, the above 3-1 valve and the above 4-1 valve are installed on the other side of the above 1 aquaculture tank, and the above 3-2 valve and the above 4-3 A valve is installed on the other side of the second aquaculture tank, and the plurality of valves includes a 5-1 valve, a 5-2 valve, a 5-3 valve, and a 5-4 valve, and the 5-4 valve is installed at the outlet, and the plurality of pumps includes a first pump, a second pump, a third pump, a fourth pump, and a fifth pump, and the first pump is connected to the 1-1 valve, and the second pump is connected between the 2-1 valve and the 5-1 valve, and the third pump is connected to the 3-1 valve and is connected between the 1-3 valve and the 1-4 valve, and the fourth pump is connected to the 3-2 valve and is connected between the 1-6 valve and the 1-7 valve, and the fifth pump is connected to the 5-4 valve, and the analysis server analyzes the tank environment data, the aquatic product data, and the external environment data to select one of a basic mode, a filtration mode, or a circulation mode, and includes the selected mode in the control signal. Transmit to the above valve gateway, andThe valve gateway changes the connection path of the plurality of pipes by opening and closing the plurality of valves according to a control signal including the selected mode, and in the basic mode, the analysis server transmits the control signal to the valve gateway to close the 2-1 valve, open the 2-2 valve, open the 2-3 valve, close the 2-4 valve, and open the 2-5 valve, the analysis server transmits the control signal to the valve gateway to close the 3-1 valve and close the 3-2 valve, the analysis server transmits the control signal to the valve gateway to open the 4-1 valve, open the 4-2 valve, open the 4-3 valve, and open the 4-4 valve, the analysis server transmits the control signal to the valve gateway to close the 5-1 valve, open the 5-2 valve, open the 5-3 valve, and open the 5-4 valve, and the analysis server transmits the control signal to the pump gateway to the first pump Driving, stopping the second pump, stopping the third pump, stopping the fourth pump, driving the fifth pump, and the analysis server selects the first aquaculture tank as the filtration tank among the plurality of aquaculture tanks in the filtration mode, and the valve gateway changes the connection path of the plurality of pipes by opening and closing the plurality of valves so that the incoming water flows through the first aquaculture tank to another aquaculture tank and is discharged to the outside, and in the filtration mode, the analysis server transmits the control signal to the valve gateway to open the 1-1 valve, close the 1-2 valve, close the 1-3 valve, open the 1-4 valve, open the 1-5 valve, and open the 1-6 valve in the plurality of pipes connected to the first aquaculture tank, the second aquaculture tank, the third aquaculture tank, and the fourth aquaculture tank.Open the above 1-7 valve, and the analysis server transmits the control signal to the valve gateway to close the 2-1 valve, open the 2-2 valve, open the 2-3 valve, close the 2-4 valve, and open the 2-5 valve; the analysis server transmits the control signal to the valve gateway to open the 3-1 valve and close the 3-2 valve; the analysis server transmits the control signal to the valve gateway to close the 4-1 valve, close the 4-2 valve, open the 4-3 valve, and open the 4-4 valve; the analysis server transmits the control signal to the valve gateway to close the 5-1 valve, close the 5-2 valve, close the 5-3 valve, and open the 5-4 valve; the analysis server transmits the control signal to the pump gateway to drive the 1 pump, stop the 2 pump, drive the 3 pump, and stop the 4 pump. A flow-through smart fish farm automated water quality management system that drives the above-mentioned fifth pump. Claim 2 A flow-through smart aquaculture farm automated water quality management system according to claim 1, wherein the analysis server measures the flow rate of the plurality of pumps, the rotational speed of the plurality of pumps, and the power consumption of the plurality of pumps through the pump gateway, determines the operation schedule of the plurality of pumps to satisfy minimum power consumption and flow rate conditions according to the relationship between the flow rate of the plurality of pumps, the rotational speed of the plurality of pumps, and the power consumption of the plurality of pumps, and includes information regarding the operation schedule of the plurality of pumps in the control signal and transmits it to the pump gateway. Claim 3 In paragraph 2, the analysis server predicts the water quality results of the plurality of aquaculture tanks using an LSTM (Long Short-Term Memory) hybrid model and a regression analysis model based on the external environment data and the tank environment data, wherein the LSTM hybrid model applies attention using the external environment data to the LSTM encoder when the LSTM decoder predicts the water quality results in an LSTM encoder-LSTM decoder combined structure based on an encoder-decoder structure, and the analysis server includes the predicted water quality results in the control signal and transmits them to the water quality treatment device, thereby forming a flow-through smart aquaculture automated water quality management system. Claim 4 In paragraph 3, the analysis server communicates with the pump gateway and the water quality treatment device using MQTT (Message Queuing Telemetry Transport) and MODBUS protocols, the pump gateway controls the plurality of pumps according to a control signal containing information regarding the operation schedule, and the water quality treatment device controls the oxygen supply according to a control signal containing the predicted water quality result, a flow-through smart aquaculture automated water quality management system. Claim 5 delete Claim 6 delete Claim 7 A flow-through smart aquaculture farm automated water quality management system according to claim 1, wherein the analysis server selects a first aquaculture tank as a sedimentation tank among the plurality of aquaculture tanks in the circulation mode and selects a second aquaculture tank as a filtration tank among the plurality of aquaculture tanks, and the valve gateway changes the connection path of the plurality of pipes by opening and closing the plurality of valves so that water flows through the first aquaculture tank to the second aquaculture tank, water flows from the second aquaculture tank to another aquaculture tank, and water flows back into the first aquaculture tank from the other aquaculture tank. Claim 8 A method of operation for an automated water quality management system for a smart aquaculture farm, comprising: a step of acquiring tank environment data regarding the environment of a plurality of aquaculture tanks through tank sensors installed in a plurality of aquaculture tanks; a step of acquiring aquatic product data regarding aquatic products through image sensors installed in the plurality of aquaculture tanks; a step of acquiring external environment data regarding seawater flowing into the plurality of aquaculture tanks through a communication device; a step of analyzing the tank environment data, the aquatic product data, and the external environment data through an analysis server to output an analysis result and generate a control signal according to the analysis result; a step of providing the analysis result as visualization data through a monitoring device; a step of controlling a plurality of valves installed in the plurality of pipes based on the control signal received from the analysis server through a valve gateway; and a step of controlling a plurality of pumps installed in the plurality of pipes based on the control signal received from the analysis server through a pump gateway. The method includes the step of treating water contained in the plurality of aquaculture tanks based on the control signal received from the analysis server through a water quality treatment device, wherein the plurality of aquaculture tanks includes a first aquaculture tank, a second aquaculture tank, a third aquaculture tank, and a fourth aquaculture tank, and the first aquaculture tank, the second aquaculture tank, the third aquaculture tank, and the fourth aquaculture tank are connected through the plurality of pipes, and the plurality of valves include a first-1 valve, a first-2 valve, a first-3 valve, a first-4 valve, a first-5 valve, a first-6 valve, and a first-7 valve, wherein the first-1 valve is installed at an inlet, the first-2 valve is connected to the first-1 valve, the first-3 valve is connected to the first-2 valve, the first-4 valve is connected to the first-3 valve, the first-5 valve is connected to the first-4 valve, and the first-6 valve is connected to the first-5 valve. The above 1-7 valve is connected to the above 1-6 valve, andThe plurality of valves includes a 2-1 valve, a 2-2 valve, a 2-3 valve, a 2-4 valve, and a 2-5 valve, wherein the 2-2 valve is installed on one side of the first aquaculture tank, the 2-3 valve is installed on one side of the second aquaculture tank, and the 2-5 valve is installed on one side of the third aquaculture tank; the plurality of valves includes a 3-1 valve, a 3-2 valve, a 4-1 valve, a 4-2 valve, a 4-3 valve, and a 4-4 valve, wherein the 3-1 valve and the 4-1 valve are installed on the other side of the first aquaculture tank, and the 3-2 valve and the 4-3 valve are installed on the other side of the second aquaculture tank; the plurality of valves includes a 5-1 valve, a 5-2 valve, a 5-3 valve, and a 5-4 valve, wherein the 5-4 valve is installed at the discharge port; and the plurality of pumps include a first pump and a second It includes a pump, a third pump, a fourth pump, and a fifth pump, wherein the first pump is connected to the first-1 valve, the second pump is connected between the second-1 valve and the fifth-1 valve, the third pump is connected to the third-1 valve and is connected between the first-3 valve and the first-4 valve, the fourth pump is connected to the third-2 valve and is connected between the first-6 valve and the first-7 valve, and the fifth pump is connected to the fifth-4 valve, and the analysis server analyzes the tank environment data, the aquatic product data, and the external environment data to select one of a basic mode, a filtration mode, or a circulation mode, includes the selected mode in the control signal, and transmits it to the valve gateway, and the valve gateway opens and closes the plurality of valves according to the control signal including the selected mode to change the connection path of the plurality of pipes, and in the basic mode, the analysis server transmits the control signal to the valve gateway to close the second-1 valve, and the second-2 Open the valve,Open the above 2-3 valve, close the above 2-4 valve, open the above 2-5 valve, and the analysis server transmits the control signal to the valve gateway to close the above 3-1 valve and close the above 3-2 valve, and the analysis server transmits the control signal to the valve gateway to open the above 4-1 valve, open the above 4-2 valve, open the above 4-3 valve, and open the above 4-4 valve, and the analysis server transmits the control signal to the valve gateway to close the above 5-1 valve, open the above 5-2 valve, open the above 5-3 valve, and open the above 5-4 valve, and the analysis server transmits the control signal to the pump gateway to drive the above 1 pump, stop the above 2 pump, stop the above 3 pump, stop the above 4 pump, and drive the above 5 pump, and the analysis server selects the above 1 aquaculture tank as the filtration tank among the above plurality of aquaculture tanks in the filtration mode, and the valve gateway The plurality of valves are opened and closed to change the connection path of the plurality of pipes so that the incoming water flows through the first aquaculture tank to another aquaculture tank and is discharged to the outside, and in the filtration mode, the analysis server transmits the control signal to the valve gateway to open the 1-1 valve, close the 1-2 valve, close the 1-3 valve, open the 1-4 valve, open the 1-5 valve, open the 1-6 valve, and open the 1-7 valve in the plurality of pipes connected to the first aquaculture tank, the second aquaculture tank, the third aquaculture tank, and the fourth aquaculture tank, and the analysis server transmits the control signal to the valve gateway to close the 2-1 valve, open the 2-2 valve, open the 2-3 valve, close the 2-4 valve, and open the 2-5 valve, and the analysis server transmits the control signal to the valve gateway,A method of operation of a flow-through smart aquaculture automated water quality management system, wherein the above-mentioned 3-1 valve is opened and the above-mentioned 3-2 valve is closed, the analysis server transmits the above-mentioned control signal to the valve gateway to close the above-mentioned 4-1 valve, close the above-mentioned 4-2 valve, open the above-mentioned 4-3 valve, open the above-mentioned 4-4 valve, the analysis server transmits the above-mentioned control signal to the valve gateway to close the above-mentioned 5-1 valve, close the above-mentioned 5-2 valve, close the above-mentioned 5-3 valve, open the above-mentioned 5-4 valve, and the analysis server transmits the above-mentioned control signal to the pump gateway to drive the above-mentioned 1st pump, stop the above-mentioned 2nd pump, drive the above-mentioned 3rd pump, stop the above-mentioned 4th pump, and drive the above-mentioned 5th pump.

Images