Ecological breeding environment intelligent regulation and control method and system based on artificial intelligence

Abstract

The invention relates to the technical field of artificial intelligence, and discloses an ecological breeding environment intelligent regulation and control method and system based on artificial intelligence, and the method comprises the steps: collecting the continuous image data of the position of a water taking grid at a preset time interval, obtaining the real-time number of target breeding objects and the active state (active or static) of each breeding object through object recognition, and obtaining the real-time number of the target breeding objects; then calculating the proportion of the cultured objects in the active state to obtain a target proportion; inputting the real-time quantity and the target proportion into an excrement estimation model to obtain an excrement yield, and finally, respectively controlling a circulating water pump to work at first and second powers higher than or lower than a preset power according to the comparison between the excrement yield and a threshold value, so as to realize the dynamic adjustment of the aquaculture water circulation replacement intensity. According to the invention, efficient utilization of water resources and energy is realized through self-adaptive regulation and control of water supply, so that fine management of an ecological breeding environment is realized.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, specifically to an intelligent control method and system for ecological aquaculture environments based on artificial intelligence. Background Technology

[0002] In the field of ecological aquaculture, the stability and cleanliness of the aquatic environment are the core elements to ensure the healthy growth of aquatic organisms. If the excrement generated during the aquaculture process is not treated in a timely manner, it can easily lead to problems such as eutrophication and decreased dissolved oxygen in the water, which in turn can cause diseases or even death of aquatic organisms, seriously affecting the aquaculture efficiency. Currently, most aquaculture farms still rely on manual inspections or timed water changes for water management. This model has obvious limitations: manual inspections make it difficult to grasp the activity status of aquatic animals and the amount of excrement produced in real time and accurately, often resulting in delayed water changes or unreasonable water change efficiency, leading to a waste of water resources and energy; traditional timed water change strategies cannot flexibly adjust the frequency of water changes according to the actual excrement production, making it difficult to achieve refined management of the ecological aquaculture environment, and thus failing to meet the urgent need for precise environmental control in ecological aquaculture. Summary of the Invention

[0003] The main objective of this invention is to provide an intelligent control method and system for ecological aquaculture environment based on artificial intelligence, aiming to solve the technical problem that existing technologies cannot meet the urgent need for precise environmental control in ecological aquaculture.

[0004] To achieve the above objectives, in a first aspect, this application provides an ecological aquaculture environment control system, comprising: Aquaculture water intake area, the aquaculture water intake area including a water supply tank and a water intake grid installed on the water supply tank; The water purification area includes a microbial decomposition tank and a physical filtration tank that are interconnected, and the microbial decomposition tank is connected to the water supply tank. An underground water storage area includes a water storage tank and a water storage pipe assembly installed inside the water storage tank. One end of the water storage tank is connected to the physical filter box, and the other end is connected to the water supply tank through a circulating water pump. A visual sensor is used to acquire image data of the target aquaculture organisms on the water intake grid; The controller is at least used to analyze the image data of the target aquaculture organism to obtain analysis results, and to perform differentiated control of the circulating water pump based on the analysis results, so as to intelligently regulate the circulation and replacement of the aquaculture water.

[0005] In one possible implementation, the physical filter box is provided with a filter screen on the side near the microbial decomposition box, and / or, the physical filter box is provided with activated carbon on the side away from the microbial decomposition box.

[0006] In one possible implementation, the water storage pipe assembly comprises multiple bamboo poles stacked along the height direction, and the length ratio of the water storage pipe assembly to the water storage tank ranges from 1 / 2 to 2 / 3.

[0007] In one possible implementation, the system further includes a water level sensor located in the water storage tank, and the processor is further configured to increase the operating power of the circulating water pump to create a flushing effect on the water storage pipe assembly when the water level in the water storage tank drops to a preset warning level and the rate of water level drop is greater than or equal to a rate threshold.

[0008] Secondly, embodiments of this application also provide an intelligent control method for ecological aquaculture environments based on artificial intelligence, applied to the ecological aquaculture environment control system as described in the first aspect, the method comprising: Multiple frames of target images are obtained by acquiring continuous image data of the water intake grid location at preset time intervals. Object recognition is performed on the multi-frame target images to obtain the real-time quantity of target aquaculture species and the activity status of each target aquaculture species, wherein the activity status includes active state and static state; The target percentage is obtained by determining the percentage of target species that are active based on their activity status. The real-time quantity of the target aquaculture species and the target proportion are input into the excrement prediction model to obtain the excrement production. If the excrement production is determined to be greater than or equal to the excrement threshold, the circulating water pump is controlled to operate at a first power to increase the circulation and replacement of the aquaculture water, wherein the first power is greater than a preset power. If the amount of excrement is determined to be less than the excrement threshold, the circulating water pump is controlled to operate at a second power to reduce the circulation and replacement of the aquaculture water, wherein the second power is less than the preset power.

[0009] In one possible implementation, the step of performing object recognition on the multi-frame target images to obtain the real-time quantity of the target aquaculture species includes: Perform object contour recognition on the multi-frame target images to obtain candidate contour lines corresponding to each frame image; If the area of ​​the region corresponding to the candidate contour line is greater than or equal to the area threshold, the object corresponding to the candidate contour line is determined as a real aquaculture product. The real-time quantity of the target cultured organism is obtained by counting the actual quantity of cultured organisms.

[0010] In one possible implementation, after obtaining the candidate contour lines corresponding to each frame image, the following is also included: Obtain the minimum area A of the target cultured organism in the historical image. min and the maximum area Amax ; According to the minimum area A min and the maximum area A max Determine the area threshold, where the area threshold T = K * (A) min +A max ) / 2, where K is a correction coefficient that matches the growth stage.

[0011] In one possible implementation, the step of performing object recognition on the multi-frame target images to obtain the activity state of the target aquaculture organism includes: The target vector is obtained by acquiring the head feature vectors of each target aquaculture species in multiple target images; Compare the similarity of target vectors representing the same farmed organism in multiple frames of target images; If the similarity of the target vector is greater than or equal to the similarity threshold, the corresponding target aquaculture species is determined to be in a static state. If the similarity between the target vector and the target vector is less than the similarity threshold, the target aquaculture species is determined to be in an active state.

[0012] In one possible implementation, the control of the circulating water pump to operate at a second power to reduce the circulation and replacement of the aquaculture water includes: During the operation of the circulating water pump at the second power, the water level status and water level change rate in the water storage tank are acquired in real time. When the water level in the storage tank drops to a preset warning level and the rate of water level drop is greater than or equal to a rate threshold, the circulating water pump is controlled to work alternately at a second power and a third power, wherein the third power is greater than the second power and the duty cycle of the third power is less than the duty cycle of the second power.

[0013] In one possible implementation, controlling the circulating water pump to operate alternately at a second power and a third power includes: The duty cycles corresponding to the second and third power are determined based on the rate of water level drop in the water tank, resulting in the first and second duty cycles. The rate of water level drop is positively correlated with the second duty cycle. The circulating water pump is controlled to operate alternately at a second power and a third power according to the first duty cycle and the second duty cycle.

[0014] Unlike existing technologies, this application provides an intelligent control method and system for ecological aquaculture environments based on artificial intelligence. This method collects continuous image data of the water intake grid location at preset time intervals, obtains the real-time quantity and activity status (active or stationary) of the target aquaculture species through object recognition, calculates the proportion of active species to obtain the target proportion, inputs the real-time quantity and target proportion into an excrement prediction model to obtain excrement production, and finally controls the circulating water pump to operate at a first and second power level higher or lower than preset power based on a comparison of excrement production with a threshold, thereby dynamically adjusting the intensity of water circulation and replacement. This application captures the quantity and activity status of aquaculture species in real time through image recognition, combined with model-predicted excrement production, enabling water circulation control to be based on accurate judgment of the actual situation at the aquaculture site, avoiding the lag of traditional manual inspections and the blindness of timed water changes. Furthermore, while ensuring water cleanliness, it avoids energy waste caused by excessive power and environmental degradation caused by insufficient power, achieving efficient utilization of water resources and energy, and thus realizing refined management of the ecological aquaculture environment. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0016] Figure 1 This is a schematic diagram of the structure of the ecological aquaculture environment control system in some embodiments of this application; Figure 2 This is a flowchart illustrating the intelligent control method for ecological aquaculture environment in some embodiments of this application; Figure 3 This is a flowchart illustrating step S200 of the intelligent regulation method for ecological aquaculture environment in some embodiments of this application; Figure 4 This is a flowchart illustrating step S200 of the intelligent regulation method for ecological aquaculture environment in some other embodiments of this application; Figure 5 This is a flowchart illustrating step S600 of the intelligent regulation method for ecological aquaculture environment in some embodiments of this application; Figure 6 This is a schematic diagram of the hardware structure of an electronic device in some embodiments of this application.

[0017] The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0019] It should be noted that all directional indications in the embodiments of the present invention are only used to explain the relative positional relationship and movement of the components in a specific posture. If the specific posture changes, the directional indications will also change accordingly.

[0020] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the term "and / or" throughout the text includes three solutions; taking A and / or B as an example, it includes technical solution A, technical solution B, and a technical solution that simultaneously satisfies A and B. Furthermore, the technical solutions of various embodiments can be combined with each other, but this must be based on the ability of a person skilled in the art to implement them. When the combination of technical solutions is contradictory or impossible to implement, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by this invention.

[0021] In the field of ecological aquaculture, the stability and cleanliness of the aquatic environment are the core elements to ensure the healthy growth of aquatic organisms. If the excrement generated during the aquaculture process is not treated in a timely manner, it can easily lead to problems such as eutrophication and decreased dissolved oxygen in the water, which in turn can cause diseases or even death of aquatic organisms, seriously affecting the aquaculture efficiency. Currently, most aquaculture farms still rely on manual inspections or scheduled water changes for water management. This approach has significant limitations: manual inspections struggle to accurately monitor the activity levels of the animals and the amount of waste produced, often leading to delayed water changes or inefficient water exchange, resulting in wasted water and energy. Furthermore, traditional scheduled water change strategies cannot flexibly adjust the frequency based on actual waste production, hindering the implementation of refined management of the ecological aquaculture environment.

[0022] To address the aforementioned problems, this application proposes an ecological aquaculture environment control system. This system is used for the intelligent control of the aquaculture water environment, such as... Figure 1As shown, the ecological aquaculture environment control system of this application includes an aquaculture water intake area 100, a water purification area 200, an underground water storage area 300, a visual sensor 400, a controller 500, and a water level sensor 600. The water flow direction of this ecological aquaculture environment control system is: underground water storage area 300 - aquaculture water intake area 100 - water purification area 200 - underground water storage area 300.

[0023] The aquaculture water intake area 100 includes a water supply tank 110 and a water intake grid 120 installed on the water supply tank 110. The water supply tank 110 is used to store water and provide drinking water for poultry such as chickens and ducks. The water intake grid 120 is a grid structure with evenly distributed water intake holes. It can serve as a support for the animals to stand upright and also allow the animals to drink water through the water intake holes, thus combining the functions of support and drinking water.

[0024] The water purification zone 200 includes an interconnected microbial decomposition tank 210 and a physical filtration tank 220. The microbial decomposition tank 210 is connected to the water supply tank 110. The microbial decomposition tank 210 is equipped with functional microorganisms that can decompose animal feces. When water containing excrement flows into the water supply tank 110, these microorganisms can efficiently degrade organic pollutants (such as livestock feces) in the water through biological metabolism. The physical filtration tank 220 has a filter screen 221 on the side closest to the microbial decomposition tank 210. The filter screen 221 can initially intercept the water after microbial decomposition, filtering out solid impurities that have not been completely decomposed. The other side of the physical filtration tank 220 away from the microbial decomposition tank 210 has activated carbon 222, which uses the porous adsorption properties of activated carbon to further adsorb small suspended particles, odor substances, and some soluble pollutants in the water. Through the synergistic effect of filter screen 221 and activated carbon 222, multi-stage physical filtration is achieved, which greatly improves the water purification efficiency and precision, and ultimately enables the treated water to meet the standards for recycling or discharge.

[0025] The underground water storage area 300 includes a water storage tank 310 and a water storage pipe assembly 320 installed in the water storage tank 310. One end of the water storage tank 310 is connected to the physical filter box 220, and the other end is connected to the water supply tank 110 through the circulating water pump 330, so that the aquaculture water intake area 100, the water purification area 200 and the underground water storage area 300 form a complete water circulation loop.

[0026] The water storage tank 310 is installed underground, which not only reduces water temperature fluctuations by utilizing the thermal insulation properties of the soil, but also directly collects and stores natural precipitation, effectively improving water resource utilization efficiency and reducing dependence on external water supply. The water storage pipe assembly 320 is composed of multiple bamboo poles stacked along the height direction. The material properties of bamboo itself have good moisture retention, which significantly reduces water evaporation loss and maintains a stable water storage volume when used as a water storage pipe.

[0027] In one embodiment, the length ratio of the water storage pipe assembly 320 to the water storage tank 310 is controlled within the range of 1 / 2 to 2 / 3, ensuring that a sufficient settling zone is reserved within the water storage tank 310. When water treated by the physical filter box 220 flows into the water storage tank 310, the settling zone allows residual fine impurities in the water to settle sufficiently, further improving the water cleanliness and providing a higher quality water source for subsequent circulating water supply.

[0028] For example, the water storage pipe assembly 330 is located in the middle of the length direction of the water storage tank 310, so that both ends of the water storage tank 310 can form independent settling zones.

[0029] The vision sensor 400 is used to acquire image data of the target aquaculture on the water intake grid 120. The vision sensor 400 can be a camera set above the water intake grid.

[0030] A water level sensor 600 is installed inside the water storage tank 310 to detect the water level status inside the tank 310, such as the water level height and the rate of water level change. The water level sensor 600 can be a float water level detector or other types of sensors, which are not limited here.

[0031] The controller 500 in this application plays a core regulatory role in the entire ecological aquaculture environment control system, specifically in two key aspects: On one hand, the controller 500 performs in-depth analysis of the image data of the target aquaculture organisms to accurately obtain key information such as the real-time quantity and activity status of each organism. Based on these analysis results, it implements differentiated power control for the circulating water pump 330. In this way, the intensity of water circulation and replacement can be dynamically adapted to the actual excretion of the aquaculture organisms. When the number of aquaculture organisms is large and the proportion of active organisms is high, the circulating water pump operates at higher power to accelerate water circulation and quickly discharge excrement from the water supply tank 110; conversely, the power is reduced to ensure water quality while reducing energy consumption, thereby constructing an intelligent water circulation control mechanism guided by the actual state of the aquaculture organisms. On the other hand, the controller 500 has the capability to monitor and handle abnormal water levels. When it detects that the water level in the storage tank 310 has dropped to a preset warning level, and the rate of water level drop is greater than or equal to a set rate threshold, it controls the power output of the circulating water pump 330 to increase. A rapid drop in water level indicates that the water storage pipe assembly 320 is about to dry out. Once the bamboo water storage pipe dries out, the fine debris remaining on its inner wall is easily dried and adhered due to moisture evaporation. If it is left until the debris dries out before cleaning, not only will the difficulty of cleaning increase significantly, but it is also very easy for incomplete cleaning to cause the inner diameter of the pipe to shrink or even become blocked, seriously affecting the efficiency of subsequent water circulation.

[0032] In this application, the controller 500 can generate a strong water flow by increasing the power of the water pump before the water storage pipe assembly 320 dries out and before debris hardens, thus providing timely and powerful flushing of the inner wall of the water storage pipe assembly 320. This pre-treatment method, which intervenes in advance, can remove potential debris to the greatest extent possible before it hardens, significantly reducing the probability of pipe blockage, effectively preventing damage to pipe function, and thus ensuring the long-term stable and efficient operation of the entire water circulation system.

[0033] Based on the same inventive objective, this application also proposes an intelligent control method for ecological aquaculture environments based on artificial intelligence, such as... Figures 1-5 As shown, the following explanation uses an ecological aquaculture environment control system as an example to illustrate the implementation of this AI-based intelligent control method for ecological aquaculture environments. It should be noted that although the logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order. Please refer to the appendix. Figure 2 The method includes the following steps S100-S600: Step S100: Obtain continuous image data of the water intake grid position at preset time intervals to obtain multiple frames of target images; It is understandable that a single frame image can only present the static form of the farmed animal at a particular instant and cannot fully reflect its movement trend; however, by continuously acquiring multiple frames of images, a dynamic sequence of the farmed animal's behavior can be formed. For example, when the farmed animal is active, its limb position and posture will show obvious changes in adjacent frames (such as head rotation, body movement, etc.); when it is stationary, the form in multiple frames remains stable. Therefore, comparative analysis of multiple target images can accurately distinguish whether the farmed animal is active or stationary, providing a reliable basis for subsequent calculation of the proportion of active states.

[0034] The preset time interval can be 1 second, 2 seconds, or other time intervals.

[0035] For example, three consecutive images are acquired at 1-second intervals to obtain three target images, which are then used as the basis for subsequent analysis.

[0036] Step S200: Perform object recognition on the multi-frame target images to obtain the real-time quantity of target aquaculture animals and the active status of each target aquaculture animal, wherein the active status includes active state and static state; It is understandable that the real-time quantity and activity level of the target aquaculture organisms on the water intake grid are directly related to the amount of excrement produced. Generally, the greater the quantity of aquaculture organisms and the higher the percentage of active organisms, the greater the amount of excrement produced per unit time. The amount of excrement produced directly determines the necessary intensity of water circulation and is the core basis for adjusting the power of the circulating water pump. The higher the amount of excrement produced, the stronger the water flow is required to accelerate its decomposition and discharge.

[0037] Therefore, after acquiring multiple target images at preset time intervals, it is necessary to perform object recognition on these target images: on the one hand, accurately count the real-time number of aquatic organisms currently at the water intake grid position to avoid deviations in basic data due to omissions or duplicate counts; on the other hand, determine the active state (active state or static state) of each aquatic organism to provide an accurate basis for subsequent calculation of the proportion of active state.

[0038] In one embodiment, such as Figure 3 As shown, step S200: performing object recognition on the multi-frame target images to obtain the real-time quantity of the target aquaculture organisms includes: S210. Perform object contour recognition on the multi-frame target images to obtain candidate contour lines corresponding to each frame image; S220. If the area of ​​the region corresponding to the candidate contour line is greater than or equal to the area threshold, the object corresponding to the candidate contour line is determined as a real aquaculture product. S230. Calculate the real-time quantity of the target cultured organism by counting the actual quantity of cultured organisms.

[0039] Specifically, each frame of the image is first processed using image recognition algorithms (such as edge detection and contour extraction) to capture the contours of all possible farmed objects and obtain corresponding candidate contour lines. In actual shooting, images may contain noise (such as falling debris or pseudo-contours formed by water ripples), and the area of ​​these pseudo-contours is usually smaller than the contour area of ​​the real farmed animals. Therefore, after obtaining the candidate contour lines for each frame, this embodiment effectively filters out non-farmed interference objects by setting a reasonable area threshold (such as based on the average body size of farmed animals like chickens and ducks), retaining only the contours of real farmed animals that match their body size characteristics. Finally, the number of real farmed animals is counted to obtain the real-time quantity of the target farmed animals.

[0040] Because the aquaculture organisms may be briefly obscured or move out of the shooting range during continuous shooting, a single frame may miss some data. The frame with the largest number of organisms usually more closely approximates the actual size of the aquaculture organisms on the current water intake grid, minimizing statistical bias caused by momentary obstruction. Therefore, the frame with the largest number of organisms among multiple target images can be taken as the real-time number of the target aquaculture organisms.

[0041] In other embodiments, the number of target aquaculture species in each frame of the target image can be calculated first, and then the average of these numbers can be taken as the real-time number.

[0042] In one embodiment, after obtaining the candidate contour lines corresponding to each frame image, the method further includes: Obtain the minimum area A of the target cultured organism in the historical image. min and the maximum area A max ; According to the minimum area A min and the maximum area A max Determine the area threshold, where the area threshold T = K * (A) min +A max ) / 2, where K is a correction coefficient that matches the growth stage.

[0043] Animals at different growth stages (such as chicks and adult chickens) exhibit significant body size differences. The smallest area A statistically analyzed in historical images is... min and the maximum area A max It can reflect the size range of the target farmed animals in the current farming scenario, providing a real and reliable reference benchmark for subsequent threshold setting, and avoiding misjudgment caused by using fixed thresholds (such as misjudging small young birds as debris, or misidentifying large debris as farmed animals). The minimum area A of the target aquaculture species in the historical image is obtained. min and the maximum area A max Then, based on the minimum area A min and the maximum area A max Determine the area threshold, where the area threshold T = K * (A) min +A max The formula (A) / 2 reflects the dynamic adaptability of the threshold setting. min +A max) / 2 represents the median value of the historical farmed animal size, serving as a basic threshold reference; while K, as a correction coefficient matched to the growth stage, can be flexibly adjusted according to the real-time growth stage of the farmed animal (such as juvenile stage, adult stage) or the shooting environment (such as light intensity, shooting distance). For example, for farmed animals in a rapid growth phase, the K value can be appropriately increased to expand the recognition range and avoid missed detections; when insufficient light causes a decrease in contour recognition accuracy, the K value can be fine-tuned to reduce false contour interference. This application embodiment determines the area threshold by combining historical data with dynamic correction, making the selection criteria for candidate outlines more in line with the actual breeding scenario. It can effectively eliminate debris (such as debris and dust) that is too small and abnormal interference (such as tools and foreign objects) that is too large, while accurately retaining the true outline that matches the characteristics of the current breeding, laying a more accurate foundation for subsequent statistical analysis of the actual number of breeding animals.

[0044] In other embodiments, such as Figure 4 As shown, step S200: performing object recognition on the multi-frame target images to obtain the activity state of the target aquaculture organism includes: S240. Obtain the head feature vector of each target aquaculture species in multiple frames of target images to obtain the target vector; S250. Compare the similarity of target vectors representing the same target aquaculture species in multiple frames of target images; S260. If the similarity of the target vector is greater than or equal to the similarity threshold, determine that the corresponding target aquaculture product is in a static state. S270. If the similarity of the target vector is less than the similarity threshold, the target aquaculture species is determined to be in an active state.

[0045] Since the head is the most active body part of poultry such as chickens and ducks, its posture changes can directly reflect the activity state of the farmed animals. Therefore, this application adopts a judgment logic that focuses on head features: First, the head feature vectors of each target farmed animal are extracted from multiple frames of target images (for example, feature vectors generated based on the contour trajectory from the root of the head to the top of the head), and the morphological features of the head are converted into quantifiable and comparable target vectors. Then, the similarity of the target vectors representing the same target farmed animal in multiple frames of images is compared. When the similarity of the target vectors is greater than or equal to a preset similarity threshold, it indicates that the head of the farmed animal has minimal posture changes in the continuous images, that is, the head is in an inactive state, and thus the corresponding target farmed animal can be determined to be in a static state. When the similarity of the target vectors is less than the similarity threshold, it indicates that there are significant posture differences in the head in the continuous images, that is, the head is in an active state, and thus the corresponding target farmed animal can be determined to be in an active state.

[0046] For a small number of objects that are difficult to identify as the same aquaculture species in multiple target images, they can be excluded from the judgment scope and not included in the judgment results of static or active states, so as to avoid such ambiguous objects interfering with the overall judgment logic and ensure the accuracy and reliability of the judgment results.

[0047] Step S300: Determine the percentage of target aquatic organisms in an active state based on the activity status of each target aquatic organism to obtain the target percentage; After obtaining the movement state (active or stationary) of each target aquaculture species through head feature vector analysis, the target proportion can be calculated through the following steps: First, count the number of all target aquaculture species determined to be active to obtain the specific number of active aquaculture species; then, calculate the ratio of this number to the real-time total number of target aquaculture species (i.e., the total number of aquaculture species on the current water intake grid determined by multi-frame image recognition), and the result is the proportion of target aquaculture species in an active state (target proportion).

[0048] Step S400: Input the real-time quantity of the target aquaculture species and the target proportion into the excrement prediction model to obtain the excrement production; The real-time quantity of target aquaculture species reflects the number of aquaculture species currently drawing water from the intake grid. Under the same activity level, the greater the number of aquaculture species, the higher the excrement output per unit time. The target percentage (the proportion of active aquaculture species) reflects the metabolic activity of the population. Generally, active aquaculture species consume energy faster, resulting in higher individual excrement output than inactive aquaculture species. Therefore, the percentage directly affects the weighting of the overall excrement output calculation.

[0049] This application embodiment uses a pre-trained excrement prediction model to predict excrement production. The excrement prediction model is a regression model (such as a neural network model or multiple linear regression model) trained on a large amount of historical aquaculture data. Internally, it includes multiple sets of mapping relationships, such as "real-time quantity - baseline production" and "target percentage - production correction coefficient." When the real-time quantity and target percentage are input, the model first calculates the baseline excrement production based on the real-time quantity, and then dynamically adjusts the baseline value based on the correction coefficient corresponding to the target percentage (for example, for every 10% increase in the activity percentage, the production prediction result increases by 8%), ultimately outputting an accurate predicted value of excrement production per unit time.

[0050] Step S500: Determine that the excrement output is greater than or equal to the excrement threshold, and control the circulating water pump to work at a first power to increase the circulation and replacement of the aquaculture water, wherein the first power is greater than a preset power; When the estimated excrement output obtained in step S400 through the excrement prediction model is greater than or equal to the preset excrement threshold, it indicates that the total amount of feces and other excrement falling into the water supply tank per unit time has reached or exceeded the upper limit that the water body can bear through self-purification and normal circulation. If water circulation is not strengthened in time, excrement will accumulate rapidly in the water supply tank, causing the concentration of harmful substances such as ammonia nitrogen and nitrite in the water to soar, destroying the living environment of the aquatic animals. At this point, the controller will trigger the corresponding control strategy: controlling the circulating water pump to operate at a first power level. This first power level is a specific power value higher than the preset baseline power, and its function is to accelerate the circulation and replacement speed of the aquaculture water by increasing the pump's operating intensity. On the one hand, this allows the water rich in excrement in the supply tank to be transported to the water purification area for treatment more quickly; on the other hand, by enhancing the water flow dynamics, it reduces the deposition of excrement at the bottom of the supply tank, reducing the difficulty of subsequent cleaning.

[0051] Step S600: Determine that the excrement output is less than the excrement threshold, and control the circulating water pump to operate at the second power to reduce the circulation and replacement of the aquaculture water, wherein the second power is less than the preset power.

[0052] When the excrement production is determined to be less than the excrement threshold by the excrement prediction model, it indicates that the total amount of feces and other excrement falling into the water supply tank per unit time is within the range that the water body can effectively bear through self-purification and regular circulation, and water quality stability can be maintained without high-intensity water circulation. At this point, the controller will activate the corresponding energy-saving control strategy: controlling the circulating water pump to operate at a secondary power level. This secondary power level is a specific power value lower than the preset baseline power. Its function is to reduce the pump's operating intensity and slow down the circulation and replacement rate of the aquaculture water. When the amount of excrement is low, the concentration of pollutants in the water increases slowly, and excessive circulation and replacement would cause unnecessary consumption of energy and water resources. Operating at the secondary power level ensures that the excrement in the water is gradually transported to the water purification zone for treatment, maintaining the water quality within acceptable limits, while also reducing equipment energy consumption by lowering the power, thus achieving energy conservation and emission reduction.

[0053] It is understandable that when the circulating water pump operates at a lower second power, the water circulation speed slows down, and the water flow impact force also weakens. Under these circumstances, fine debris suspended in the water storage tank (such as incompletely decomposed excrement residue, feed scraps, etc.) is likely to gradually accumulate inside the water storage pipe assembly with the slow water flow; at the same time, if the water level in the water storage tank drops due to evaporation or consumption, the slower-flowing water cannot flush the inner wall of the water storage pipe assembly in time, and the accumulated debris will gradually dry and adhere due to the reduction of moisture. These dried-up debris can reduce the effective flow cross-section of the water storage pipe components, creating a significant obstruction effect. This can lead to a decrease in water circulation efficiency, making it impossible to meet the water supply needs during peak aquaculture periods. In severe cases, it can cause partial blockage of the pipes, disrupting the continuity of the entire water circulation system and posing a potential threat to the stability of subsequent aquaculture water supply.

[0054] Based on this, in one embodiment, such as Figure 5 As shown, step S600: controlling the circulating water pump to operate at a second power to reduce the circulation and replacement of the aquaculture water, includes: S610. During the operation of the circulating water pump at the second power, the water level status and water level change rate in the water storage tank are acquired in real time. S620. When the water level in the water storage tank drops to a preset warning level and the rate of water level drop is greater than or equal to a rate threshold, the circulating water pump is controlled to work alternately at a second power and a third power, wherein the third power is greater than the second power and the duty cycle of the third power is less than the duty cycle of the second power.

[0055] Specifically, by continuously collecting data through devices such as water level sensors, abnormal changes in the water level of the storage tank can be detected in a timely manner. When the water level drops to a preset warning level (such as the highest point of the water storage pipe assembly) and the rate of drop is rapid, it indicates that the water storage pipe assembly is at risk of drying out, and an intervention mechanism needs to be activated. At this time, the circulating water pump is controlled to operate alternately at the second and third power levels. The third power is greater than the second power, which can generate a strong water flow through short-term high-power operation to flush away the accumulated debris on the inner wall of the water storage pipe assembly and prevent it from drying out; while the duty cycle of the third power is less than that of the second power (i.e., short high-power operation time and long low-power operation time), so that while completing the flushing task, the energy-saving characteristics of low-power operation are maintained to the maximum extent.

[0056] In one embodiment, the duty cycles corresponding to the second power and the third power can be determined first based on the rate of water level drop in the water tank to obtain the first duty cycle and the second duty cycle, wherein the rate of water level drop is positively correlated with the second duty cycle; then the circulating water pump is controlled to work alternately with the second power and the third power according to the first duty cycle and the second duty cycle.

[0057] Specifically, a faster water level drop indicates a higher risk of the water storage pipe assembly drying out and a greater likelihood of debris buildup. Therefore, the second duty cycle needs to be increased, allowing the circulating water pump to operate at the third power level for a longer period, reducing debris accumulation through more frequent, strong water flow flushing. Conversely, a slower water level drop indicates a lower risk of drying out, allowing the second duty cycle to be reduced, decreasing the high-power operation time and saving energy. Based on this logic, a mapping table between the water level drop rate and the second duty cycle can be pre-trained, and then the second duty cycle can be obtained by querying this mapping table based on the water level drop rate in the storage tank.

[0058] Next, based on the calculated first and second duty cycles, the circulating water pump is controlled to alternate between the second and third power levels. For example, if the first duty cycle is 70% and the second duty cycle is 30%, then within one operating cycle, the pump runs at the second power at low speed for 70% of the time and at the third power at high speed for 30% of the time. This dynamic adjustment mechanism allows for precise matching of flushing intensity and risk level, ensuring the cleanliness of the water storage pipe components while further optimizing energy utilization efficiency, making the alternating working mode more in line with actual operating conditions.

[0059] Based on this, this application provides an intelligent control method and system for ecological aquaculture environment based on artificial intelligence. This method collects continuous image data of the water intake grid location at preset time intervals, obtains the real-time quantity and activity status (active or stationary) of the target aquaculture species through object recognition, calculates the proportion of active aquaculture species to obtain the target proportion, inputs the real-time quantity and target proportion into an excrement prediction model to obtain excrement production, and finally controls the circulating water pump to operate at a first and second power higher or lower than preset power based on a comparison of excrement production with a threshold, thereby dynamically adjusting the intensity of water circulation replacement. This application captures the quantity and activity status of aquaculture species in real time through image recognition, and combines this with model-predicted excrement production, enabling water circulation control to be based on accurate judgment of the actual situation at the aquaculture site, avoiding the lag of traditional manual inspections and the blindness of timed water changes. Furthermore, while ensuring water cleanliness, it avoids energy waste caused by excessive power and environmental degradation caused by insufficient power, achieving efficient utilization of water resources and energy, thus realizing refined management of the ecological aquaculture environment.

[0060] like Figure 6 As shown, Figure 6 The diagram below illustrates the hardware structure of an electronic device in some embodiments of this application. The electronic device provided in the embodiments of this application includes a memory 1000 and a processor 2000. The memory 1000 is used to store computer-readable instructions, and the processor 2000 is used to call the computer-readable instructions to execute the intelligent control method for ecological aquaculture environment as described above.

[0061] The processor 2000 provides computing and control capabilities to control electronic devices to perform corresponding tasks, such as controlling the electronic devices to perform the intelligent regulation method for ecological aquaculture environment in any of the above method embodiments. The method includes: acquiring continuous image data of the water intake grid location at preset time intervals to obtain multiple frames of target images; performing object recognition on the multiple frames of target images to obtain the real-time quantity of target aquaculture organisms and the activity status of each target aquaculture organism, wherein the activity status includes an active state and a static state; determining the proportion of target aquaculture organisms in an active state based on the activity status of each target aquaculture organism to obtain a target proportion; inputting the real-time quantity of target aquaculture organisms and the target proportion into an excrement prediction model to obtain excrement production; determining that the excrement production is greater than or equal to an excrement threshold, controlling the circulating water pump to operate at a first power to increase the circulation and replacement of the aquaculture water, wherein the first power is greater than a preset power; determining that the excrement production is less than the excrement threshold, controlling the circulating water pump to operate at a second power to reduce the circulation and replacement of the aquaculture water, wherein the second power is less than a preset power.

[0062] The processor 2000 can be a general-purpose processor, including a central processing unit, a network processor, a hardware chip, or any combination thereof; it can also be a digital signal processor, an application-specific integrated circuit, a programmable logic device, or a combination thereof. The aforementioned PLD can be a complex programmable logic device, a field-programmable gate array, a general-purpose array logic, or any combination thereof.

[0063] The memory 1000, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs, non-transitory computer-executable programs, and modules, such as the program instructions / modules corresponding to the intelligent control method for the ecological aquaculture environment in the embodiments of this application. The processor 2000 can implement the intelligent control method for the ecological aquaculture environment in any of the above method embodiments by running the non-transitory software programs, instructions, and modules stored in the memory 1000.

[0064] Specifically, memory 1000 may include volatile memory, such as random access memory; memory 1000 may also include non-volatile memory, such as read-only memory, flash memory, hard disk or solid-state disk or other non-transitory solid-state storage devices; memory 1000 may also include combinations of the above types of memory.

[0065] In summary, the electronic device of this application adopts the technical solution of any of the above-described embodiments of the intelligent control method for ecological aquaculture environment. Therefore, it has at least the beneficial effects brought about by the technical solutions of the above embodiments, which will not be elaborated further here.

[0066] This application also provides a computer-readable storage medium, such as a memory including program code, which can be executed by a processor to complete the intelligent control method for ecological aquaculture environment described in the above embodiments. For example, the computer-readable storage medium may be a read-only memory, a random access memory, a read-only optical disc, a magnetic tape, a floppy disk, and an optical data storage device, etc.

[0067] This application also provides a computer program product, which includes one or more lines of program code stored in a computer-readable storage medium. The processor of the early warning system reads the program code from the computer-readable storage medium and executes the program code to complete the steps of the intelligent control method for the ecological aquaculture environment provided in the above embodiments.

[0068] Those skilled in the art will understand that all or part of the steps of the above embodiments can be implemented by hardware, or by a program or program code related to hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a disk, or an optical disk.

[0069] It should be noted that the device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0070] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software and a general-purpose hardware platform, or of course, using hardware. Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0071] The above description is merely a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural transformations made using the contents of the present invention's specification and drawings under the inventive concept of the present invention, or direct / indirect applications in other related technical fields, are included within the patent protection scope of the present invention.

Claims

1. An ecological aquaculture environment regulating system for regulating the environment of an aquaculture water body, characterized in that, The system comprises: a breeding water intake area, comprising a water supply tank and a water intake grid arranged on the water supply tank; a water quality purification area, comprising a microorganism decomposition tank and a physical filtration tank in communication with each other, the microorganism decomposition tank being in communication with the water supply tank; an underground water storage area, comprising a water storage tank and a water storage pipe assembly arranged in the water storage tank, one end of the water storage tank being in communication with the physical filtration tank, and the other end being in communication with the water supply tank through a circulating water pump; a visual sensor configured to acquire image data of target aquatic animals on the water intake grid; a controller configured to analyze the image data of the target aquatic animals to obtain an analysis result, and to differentially control the circulating water pump based on the analysis result, so as to intelligently regulate the circulation replacement of the breeding water.

2. The controlled system of claim 1, wherein, The physical filtration tank is provided with a filter screen on the side close to the microorganism decomposition tank, and / or is provided with activated carbon on the side away from the microorganism decomposition tank.

3. The controlled system of claim 1, wherein, The water storage pipe assembly comprises a plurality of bamboo pipes stacked along the height direction, and the length ratio of the water storage pipe assembly to the water storage tank is in the range of 1 / 2-2 / 3.

4. The controlled system of claim 1, wherein, The system further comprises a water level sensor arranged in the water storage tank, and the processor is further configured to increase the working power of the circulating water pump to form a scouring effect on the water storage pipe assembly when the water level of the water storage tank decreases to a preset warning water level and the water level decrease rate is greater than or equal to a rate threshold.

5. An artificial intelligence-based ecological breeding environment intelligent regulation method, characterized in that, The method is applied to the ecological breeding environment regulation system according to any one of claims 1-4, and the method comprises: acquiring image data of the water intake grid position continuously to obtain a plurality of target images at a preset time interval; performing object recognition on the plurality of target images to obtain a real-time number of target aquatic animals and an active state of each target aquatic animal, wherein the active state comprises an active state and a static state; determining a target proportion of the target aquatic animals in the active state based on the active state of each target aquatic animal to obtain a target proportion; inputting the real-time number of target aquatic animals and the target proportion into a excrement estimation model to obtain an excrement yield; determining that the excrement yield is greater than or equal to an excrement threshold, and controlling the circulating water pump to work at a first power to increase the circulation replacement of the breeding water, the first power being greater than a preset power; determining that the excrement yield is less than the excrement threshold, and controlling the circulating water pump to work at a second power to reduce the circulation replacement of the breeding water, the second power being less than the preset power.

6. The artificial intelligence-based ecological breeding environment intelligent regulation method according to claim 5, characterized in that, The object recognition on the plurality of target images to obtain the real-time number of target aquatic animals comprises: performing object contour recognition on the plurality of target images to obtain candidate contour lines corresponding to each frame of image; determining an object corresponding to the candidate contour line as a real aquatic animal when an area corresponding to the candidate contour line is greater than or equal to an area threshold; counting the number of real aquatic animals to obtain the real-time number of target aquatic animals.

7. The artificial intelligence-based ecological breeding environment intelligent regulation method according to claim 6, characterized in that, After obtaining the candidate contour lines corresponding to each frame of image, the method further comprises: Obtaining the minimum area A of the target aquaculture animal in the historical image min and the maximum area A max ; According to the minimum area A min and maximum area A max Determine the area threshold, where the area threshold T = K * (A min +A max ) / 2, K is the correction coefficient matched with the growth stage. 8.The AI-based ecological breeding environment intelligent regulation method of claim 5, wherein, The object recognition on the plurality of target images to obtain the active state of target aquatic animals comprises: Obtaining a head feature vector of each target aquaculture in multiple target images to obtain a target vector; Comparing the similarity of the target vectors representing the same target aquaculture in multiple target images; Determining that the target vector similarity is greater than or equal to the similarity threshold to determine that the corresponding target aquaculture is in a static state; Determining that the target vector similarity is less than the similarity threshold to determine that the target aquaculture is in a dynamic state. 9.The AI-based ecological breeding environment intelligent regulation method of claim 5, wherein, The control of the circulating water pump to work at the second power to reduce the circulating replacement of the aquaculture water body includes: During the process of the circulating water pump working at the second power, the water level state and the water level change rate in the water storage tank are obtained in real time; When the water level in the water storage tank drops to a preset warning water level and the water level drop rate is greater than or equal to the rate threshold, the circulating water pump is controlled to work alternately at the second power and the third power, wherein the third power is greater than the second power, and the duty cycle of the third power is less than the duty cycle of the second power. 10.The AI-based ecological breeding environment intelligent regulation method of claim 9, wherein, The control of the circulating water pump to work alternately at the second power and the third power includes: According to the water level drop rate in the water storage tank, the duty cycles corresponding to the second power and the third power are determined to obtain a first duty cycle and a second duty cycle, wherein the water level drop rate and the second duty cycle are in a positive correlation; According to the first duty cycle and the second duty cycle, the circulating water pump is controlled to work alternately at the second power and the third power.

Images