An intelligent feeding method and system for recirculating aquaculture fish based on repeated game

Abstract

The application discloses a kind of based on repeated game's circulating water aquaculture fish intelligent feeding method and system, the method includes constructing repeated game, based on the Nash equilibrium strategy of repeated game, determine the income relationship between different strategies in research phase, finally according to the internal connection of the energy intake and consumption in starvation degree and fish feeding behavior, quantification starvation degree and determine feeding amount.The repeated game method of the application is simple and effective to determine the feeding amount to promote fish cooperative swimming feeding behavior, can effectively reduce the feed waste caused by determining the feeding amount by the total amount of fish combined with the experience of aquaculture, can effectively promote the cooperative swimming feeding behavior of fish individuals while fully meeting the energy required for fish growth, avoid deterioration of aquaculture water quality, improve the efficiency of aquaculture, meet the development needs of modern aquaculture.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of energy consumption analysis of fish swimming behavior and feeding amount method of recirculating aquaculture system, and particularly relates to a recirculating aquaculture fish intelligent feeding method and system based on repeated game. BACKGROUND

[0002] The recirculating aquaculture system (RAS) is a new type of aquaculture mode, which can handle the wastewater generated in the culture pond through a series of water treatment units and then recycle it for use. It can solve the problem of low water resource utilization rate, and can save 90%-99% of water resources compared with the traditional aquaculture system. It can also provide a stable, reliable, comfortable and high-quality living environment for the cultured organisms, and provides favorable conditions for high-density culture. Such a system is also considered as the inevitable development trend of future fisheries. In the recirculating aquaculture mode, in addition to the initial equipment cost, feed is the main part of the cost of aquaculture. In the existing feeding method, sufficient feed is usually fed to the fish to ensure that the growth is not affected. Although this feeding method can ensure that the fish can intake sufficient energy, it will cause feed residues, which not only increases unnecessary feed costs, but also increases the water treatment cost of the recirculating water system. In addition, sufficient food will also destroy the swimming cooperation between individuals in the fish group. From the viewpoint of hydrodynamics, fish swimming in groups can save the energy consumption of individual fish, but when the food is sufficient, the individuals in the fish group have a food source to make up for the excess energy consumption, which will cause some individuals to swim away from the group, resulting in unnecessary swimming energy consumption, reducing the feed conversion rate of aquaculture, and reducing the yield of aquaculture.

[0003] Through observation, it is found that hungry fish individuals reduce their unnecessary swimming energy consumption by increasing swimming cooperation, and after a period of feeding, individuals that swim alone will appear in the fish group. This shows that a certain degree of hunger will promote the cooperation between fish individuals, and after a certain amount of food, the fish individuals have the ability to pay for the extra swimming energy consumption, at which time the individual may swim alone. Therefore, in order to avoid the destruction of the cooperative swimming of the fish group and save feed, we need to determine this degree of hunger, and by feeding no more than the amount of feed at each feeding, the fish group retains a certain degree of hunger to promote the cooperative swimming of the fish group.

[0004] In summary, the present application proposes a recirculating aquaculture fish intelligent feeding method based on repeated game. By feeding a sufficient amount of feed in multiple times, a limited repeated game is constructed to determine the appropriate degree of hunger, and the obtained degree of hunger is used to further determine the scientific feed feeding amount. SUMMARY

[0005] The application aims to provide an intelligent feeding method and system for fish in recirculating aquaculture based on repeated game, and provide good technical support for feeding in recirculating aquaculture.

[0006] The technical solution adopted by the application is as follows:

[0007] An intelligent feeding method for fish in recirculating aquaculture based on repeated game, the method is to utilize the relationship between the hunger degree of fish group and the feeding swimming strategy, the feeding swimming strategy is cooperative swimming feeding and uncooperative swimming feeding, based on repeated game of feeding swimming, the income relationship between different strategies of fish group when uncooperative swimming feeding appears in fish group is determined, the hunger degree is quantified by combining the energy intake and consumption of fish group, and finally the subsequent feeding amount is determined according to the hunger degree H which fish group wants to keep and the satiation feeding amount M.

[0008] Further, the method comprises the following:

[0009] Put the fish group in a recirculating aquaculture pond which has not fed for more than 24 hours, the number of fish group individuals is N, the body size is basically consistent, the satiation feeding amount M of the fish group is determined, the feeding amount each time is m, wherein M=mT, T is the number of repeated feeding; uncooperative swimming feeding appears in the fish group at the n+1th feeding, that is, the total income of fish group using this strategy is greater than the total income of fish group using cooperative swimming feeding strategy for the first n+1 times, based on this, it can be determined that the income of fish individual using uncooperative swimming feeding strategy at the n+1th feeding is greater than the income of fish individual using cooperative swimming feeding strategy.

[0010] Further, the energy income of feeding and the swimming energy consumption are used to quantify the income, and according to the income relationship of two different strategies of fish individual at the n+1th feeding, the maximum energy consumption and the ability to pay additional swimming energy consumption that can be borne by fish individual are quantitatively determined, and the reciprocal of the two is subtracted by 1, that is, the hunger degree, which is used as the hunger degree of fish group.

[0011] Further, the oxygen consumption required for fish group to complete feeding once under the condition of overall cooperative swimming feeding is Wherein the conversion coefficient between energy consumption and oxygen consumption is F1, and the consumption of individuals in fish group to complete cooperative feeding once is:

[0012]

[0013] The energy income GAN obtained is related to the feed intake of individual, and the correlation coefficient is F2, so the energy income of individual in fish group to complete cooperative feeding once is:

[0014] GAN=F2×m / N

[0015] Therefore, the income of fish individual using cooperative swimming feeding strategy at the n+1th feeding is:

[0016]

[0017] When it is detected that in the n+1th feeding process, the individual in the fish population adopts the uncooperative swimming feeding strategy, the feeding amount of any one of the individuals in the nth feeding process and the n+1th feeding process is determined by the high-definition camera, the feeding benefit improvement ratio S is determined, that is, the ratio of the feeding amount in the n+1th feeding process to the feeding amount in the nth feeding process, at this time the energy benefit of the individual completing once uncooperative feeding is:

[0018] GAN′=S×F2×m / N

[0019] Let the oxygen consumption of the individual in the n+1th feeding process be Then the consumption of the individual completing once uncooperative feeding at this time is:

[0020]

[0021] Then, the benefit of the n+1th feeding fish individual under the uncooperative swimming feeding strategy is:

[0022]

[0023] From the benefit relationship D≥C, we get:

[0024]

[0025] Then the desired hunger degree H of the fish population is:

[0026]

[0027] Further, after determining the desired hunger degree H of the fish population, the feeding amount M P of each feeding can be determined as:

[0028] M P =H×M

[0029] An intelligent feeding method for circular water aquaculture fish based on repeated game, including directly measuring the corresponding parameters according to the calculation formula of the appropriate hunger degree H, and then feeding to the satiety feeding amount M P .

[0030] An intelligent feeding system for circular water aquaculture fish based on repeated game, including a circular water aquaculture system, a high-definition camera, an LED lamp, a computer, an oxygen consumption detection device, a feeding device and a PLC controller.

[0031] The high-definition camera is installed above the recirculating aquaculture system and is connected with a computer, so that the fish group behavior can be monitored, LED lamps are used for light supplement of the high-definition camera, and the output end of the PLC controller is connected with the feeding device.

[0032] The inventive principle of the present application is:

[0033] Feeding behavior and swimming behavior are basic behavior patterns of fish, which are related to energy intake and energy consumption of fish respectively, in addition, the hunger degree of fish also affects the selection of fish on feeding swimming strategy, when fish is in an empty stomach state, it cannot bear the additional swimming energy consumption of non-collaborative swimming feeding strategy, and when the hunger degree decreases to a certain range, fish individuals will choose different feeding swimming strategies according to their own preferences, the internal relationship between hunger degree and feeding swimming strategy is utilized to define the repeated game of feeding swimming, and the hunger degree is quantified in combination with fish group behavior. A new feeding method is formulated in combination with the hunger degree H and the satiety feeding amount M.

[0034] The intelligent feeding method for recirculating aquaculture fish based on repeated game of the present application fully utilizes the behavior information of the cultured fish to reflect the selection of fish individuals between collaborative swimming and feeding strategies, gets rid of the feed waste caused by traditional feeding, realizes the accurate control of the feeding amount in recirculating aquaculture, and effectively improves the production efficiency of recirculating aquaculture while meeting the energy supply required by the growth of the cultured fish group. BRIEF DESCRIPTION OF DRAWINGS

[0035] Figure 1 It is a schematic diagram of limited repeated game based on multiple feeding. DETAILED DESCRIPTION

[0036] The technical solutions of the present application will be further described in detail below in combination with the drawings and specific embodiments.

[0037] The embodiment of the present application provides an intelligent feeding system for recirculating aquaculture fish based on repeated game, which comprises a recirculating aquaculture system, a high-definition camera, LED lamps, a computer, an oxygen consumption detection device, a feeding device and a PLC controller.

[0038] The high-definition camera is installed above the recirculating aquaculture system and is connected with a computer, so that the fish group behavior can be monitored, LED lamps are used for light supplement of the high-definition camera, and the output end of the PLC controller is connected with the feeding device.

[0039] The system executes an intelligent feeding method for recirculating aquaculture fish based on repeated game, the intelligent feeding method determines the satiety feeding amount of the cultured fish group by analyzing the collaborative swimming of the cultured objects, in an embodiment of the present application, the method specifically comprises the following steps:

[0040] 1) Put the fish population (ensure the fish body shape is basically the same, the number of individuals N) which has not been fed for more than 24 hours (i.e. in the fasting state) into the recirculating aquaculture pond, and ensure that the high-definition camera above the recirculating aquaculture pond can transmit image information to the computer in real time;

[0041] 2) Determine the satiation feeding amount M of the fish population by the existing method, repeat T times feeding, each feeding amount is m (where M = mT). Use granular floating feed when feeding, basically no time interval is needed for feeding, only need to observe after each round of feeding when the high-definition camera detects that there is no feed residue in the aquaculture pond, then start the next feeding; this process is to divide the normal feeding according to experience into T times, the purpose is to determine the income relationship between different strategies of the fish population based on the repeated game of feeding swimming, and obtain the suitable hunger degree H that the human being wants the fish population to maintain, after determining H, more accurate feeding can be carried out.

[0042] 3) At this time, the feeding strategy of each individual fish in the fish population at each feeding stage faces a game G given by the human being, in which the subject (fish) has two feeding strategies to choose from, namely cooperative swimming feeding and non-cooperative swimming feeding, where cooperative swimming and non-cooperative swimming are two swimming states of the fish population, which are terms in the art, cooperative swimming is also called swarm swimming, in this application, the feeding behavior and swimming behavior of the fish population are combined together and named as cooperative swimming feeding and non-cooperative swimming feeding; the income of cooperative swimming feeding is denoted as C, and the income of non-cooperative swimming feeding is denoted as D. The calculation method of the two incomes is consistent, which is the sum of the energy income of feeding and the energy consumption of swimming. Since the swimming behavior of the fish population changes during feeding, if the entire feeding process is regarded as a process, the income cannot be accurately analyzed, therefore, in this application, the entire feeding process is divided into T times by using the theory of repeated game, and when analyzing the income, the overall repeated game should be considered instead of only a single game.

[0043] 4) For the repeated game G, we denote the whole as G(T), which represents the finite repeated game. Through high-definition camera observation, it is found that at the nth feeding, all fish individuals still adopt cooperative swimming feeding, and at the n+1th feeding, some fish individuals change their strategies and start to use non-cooperative strategies until the completion of the entire feeding stage (n < n+1 ≤ T).

[0044] 5) At this time, we define two other repeated games G(n), G(n+1). Let R represent the total income of the repeated game, at the nth stage, if the cooperative strategy is adopted, the income is C n , if the non-cooperative strategy is adopted, the income is D n , then the income of the Nash equilibrium strategy of the repeated game G(n) is as follows:

[0045]

[0046] where δ is the discount factor, which can be seen as the patience of the participants in game theory, and the value is in [0, 1], the greater the patience of the participants, if it is equal to 0, it means that the participants have no patience at all, in the process of this study, the feeding amount and feeding time of each round are controlled to keep the discount factor stable, which does not affect the subsequent process.

[0047] 6) For another repeated game G(n+1), the total revenue has two kinds:

[0048]

[0049]

[0050] According to the observed behavior of fish, we find that R n+1 ≥R′ n+1 , that is, D n+1 ≥C n+1 ;

[0051] 7) The oxygen consumption of the fish group to complete a stage of feeding in the case of overall cooperative swimming feeding is where the conversion coefficient between energy consumption and oxygen consumption is F1, then the consumption of an individual in the fish group to complete a cooperative feeding is:

[0052]

[0053] The energy gain GAN is related to the feed intake of the individual, and the correlation coefficient is set as F2, then the energy gain of an individual in the fish group to complete a cooperative feeding is:

[0054] GAN=F2×m / N

[0055]

[0056] 8) When it is detected that an individual in the fish group adopts a non-cooperative swimming feeding strategy in the n+1th round of feeding, the feeding amount of the individual in the nth round of feeding and the feeding amount in the n+1th round of feeding are determined through the high-definition camera, and the feeding revenue improvement ratio S (the feeding amount in the n+1th round of feeding compared with the feeding amount in the nth round of feeding) is determined, at this time, the energy gain of the individual to complete a non-cooperative feeding is:

[0057] GAN′=S×F2×m / N

[0058] In the above process, there may be multiple individuals with non-cooperative swimming feeding strategies. In this study, the behavior of individuals is considered as a common extension to the group, and only one individual is concerned.

[0059] When some individuals in the fish school use non-cooperative swimming feeding strategies, the swimming routes of other individuals may also be affected, leading to more oxygen consumption. The oxygen consumption of individuals with non-cooperative swimming feeding strategies cannot be accurately measured based on the total oxygen consumption of the fish school, so the oxygen consumption of non-cooperative feeding individuals in the n+1th feeding process is The consumption of one non-cooperative feeding by the individual is:

[0060]

[0061]

[0062] 9) Combine the income quantified by the fish school in 7) and 8) with the income relationship determined in 6), and through derivation, we can get:

[0063]

[0064] The maximum energy consumption that an individual can bear is determined by the hunger level H of the current feeding stage. The hunger level H is between 0 and 1, and the greater the value, the hungrier the fish is. When H is 1, the fish is in an empty stomach state, and when H is 0, the fish is in a full stomach state. That is, as the hunger level (H) decreases, the energy consumption of fish individuals in swimming increases. According to the maximum expenditure of oxygen consumption of individuals in non-cooperative swimming feeding, subtract the consumption required for cooperative swimming feeding The maximum value of the additional expenditure is determined by 1 minus the ratio of the ability of the fish individual to pay for the additional swimming energy consumption to the maximum energy consumption that the fish individual can bear. The hunger level of the individual at this time is determined by 1 minus the ratio of the ability of the fish individual to pay for the additional swimming energy consumption to the maximum energy consumption that the fish individual can bear. The hunger level of the individual at this time is extended to the group, i.e. the appropriate hunger level H that the fish school wants to maintain is:

[0065]

[0066] 10) At the nth feeding, fish individuals still use cooperative swimming feeding, and at the n+1th feeding, some fish individuals change their strategy and start using non-cooperative strategy. At the nth feeding, we can roughly describe the hunger level of the fish school at this time as:

[0067]

[0068] At the n+1th feeding, we can roughly describe the hunger level of the fish school at this time as:

[0069]

[0070] According to the research idea of the present application, it can be determined that the amount of nm feed has been fed and the amount of (n+1) m feed is being fed according to the current feeding stage, so when the non-cooperative swimming feeding behavior occurs, the feeding amount is greater than or equal to nm and less than or equal to (n+1) m, then the hunger degree of the research object in the corresponding research period is between the two, that is:

[0071]

[0072] By this method, the accuracy of taking the hunger degree H as the appropriate hunger degree that the fish population wants to maintain is further verified. The subsequent feeding method can be determined according to the hunger degree H, and the satiation feeding amount M can be determined, and the specific feeding amount M P is:

[0073]

[0074] For the same batch of farmed fish, only the parameters in the formula for determining the appropriate hunger degree H in the present application need to be measured to determine H once, and then each feeding is performed according to Mp until the satiation feeding amount is reached. In the process of determining H:

[0075] S is the feeding amount of the research object individual in the n+1th round of feeding divided by the feeding amount in the nth round (or the average feeding amount of the previous n rounds can be used instead).

[0076] F1 is the conversion relationship between oxygen consumption and energy consumption, which is related to the flow rate in the recirculating water system, the type of fish, the mass of the fish and the temperature, and can be calculated and measured according to the existing method.

[0077] F2 is the relationship between the amount of feed intake and energy absorption, which is related to the composition of the feed and the absorption capacity of the fish, and can be measured according to the existing method.

[0078] The above disclosure is only a specific embodiment of the present application, but the present application is not limited thereto. For those skilled in the art, any modification made without departing from the present application shall be deemed to fall within the scope of the present application.

Claims

1. A method for intelligent feeding of cyclically water-bred fish based on repeated game, characterized in that, The method utilizes the relationship between the hunger level of a fish population and its feeding and swimming strategies, namely cooperative and non-cooperative swimming for feeding. Based on repeated game theory of feeding and swimming, the payoff relationship between different strategies of the fish population when non-cooperative swimming for feeding occurs is determined. The hunger level is quantified by combining the energy intake and expenditure of the fish population. Finally, the subsequent feeding amount is determined based on the desired hunger level H of the fish population and the amount of food given M. The method includes the following: A group of fish, N in number and of similar size, that have not eaten for more than 24 hours are placed in a recirculating aquaculture pond. The satiated feeding amount M is determined, and the amount fed each time is m, where M = mT and T is the number of feeding repetitions. On the (n+1)th feeding, some fish begin to swim uncooperatively to feed. This means that the total benefit of the fish using this strategy is greater than the total benefit of the fish using the cooperative swimming feeding strategy in the previous n+1 feedings. Based on this, it can be determined that the benefit of the fish using the uncooperative swimming feeding strategy on the (n+1)th feeding is greater than the benefit of the cooperative swimming feeding strategy. The energy gains from feeding and the energy consumption from swimming are used to quantify the gains. Based on the relationship between the gains of the two different strategies for individual fish at the (n+1)th feeding, the maximum energy consumption that an individual fish can bear and its ability to bear additional swimming energy consumption are quantified. The hunger level is obtained by subtracting the inverse ratio of the two from 1.

2. The intelligent feeding method for recirculating aquaculture fish based on repeated game theory according to claim 1, characterized in that, The oxygen consumption of a fish school to complete a feeding in the case of swimming feeding in whole cooperation is measured as where the conversion coefficient between energy consumption and oxygen consumption is The energy consumption of an individual in the fish school to complete a cooperative feeding is : ； The energy gain GANobtained is related to the amount of feed ingested by the individual, and the correlation coefficient between energy absorption and feed intake is The energy gain GANobtained by an individual in a school completing one cooperative meal is then : ； Then, the benefit of the fish individual cooperative foraging strategy under the n+1th feeding is: ； When detected at the time During round feeding, if an individual in the fish school adopts a non-cooperative swimming and feeding strategy, a high-definition camera can be used to identify any one of these individuals in the [number of feedings]. Feed intake during round feeding and the first The amount of food consumed in each feeding cycle determines the percentage increase in feeding benefit, S, which is the amount of food consumed in the first feeding cycle. Feed intake during round feeding and the first The ratio of food intake during round feeding, representing the energy gain of the individual during one uncooperative feeding cycle. for: ； Let the individual in the first The oxygen consumption during the round feeding process is Then, at this point, the individual has completed one cycle of uncooperative feeding. for: ； Then, the profit for the (n+1)th feeding session is calculated under the strategy of individual fish not cooperating in swimming and feeding. for: ； From the payoff relationship D≥C, we get: ； The desired level of hunger H for the fish population is: 。 3. The intelligent feeding method for recirculating aquaculture fish based on repeated game theory according to claim 1, characterized in that, After determining the desired hunger level H for the fish population, the amount of food given at each subsequent feeding will be determined. It can be determined as: 。 4. A smart feeding system for recirculating aquaculture fish based on repeated game theory, characterized in that, Includes recirculating aquaculture system, high-definition camera, LED lights, computer, oxygen consumption detection device, feeding equipment and PLC controller; A high-definition camera is installed above the recirculating aquaculture system and connected to a computer to ensure monitoring of fish behavior. LED lights are used to supplement the lighting of the high-definition camera. The output of the PLC controller is connected to the feeding equipment. The computer determines the feeding amount according to the method described in any one of claims 1-3 and transmits it to the PLC controller.

Images