阅读说明：本技术 一种对虾生态养殖方法 (Ecological prawn breeding method ) 是由 熊建华 张彬 黎铭 韦信贤 童桂香 杨艳 林勇 于 2021-01-15 设计创作，主要内容包括：本发明公开了一种对虾生态养殖方法,包括以下步骤：S1:将养殖区域进行分割形成若干个生态单元,生态单元之间采用挡土墙分隔,生态单元之间采用管道相连,管道通过控制阀控制且单向输出；S2:在每个生态单元中放入对虾、与对虾组成生态的水草和藻类,并设置有水质传感器、生态检测器；S3:进行水生态的动态调整；S4:对不满足生态平衡的生态单元进行饲料和水生态处理。本发明将养殖区域进行分割形成若干个生态单元,通过强化学习算法使各个生态单元之间进行相互调整达到动态的生态平衡,再对无法达到生态平衡的生态单元进行水生态处理,可以极大的减少水生态处理时所需的药物和人力成本,提高养殖效率。(The invention discloses an ecological prawn breeding method, which comprises the following steps: s1, dividing the cultivation area into a plurality of ecological units, wherein the ecological units are separated by retaining walls and connected by pipelines controlled by control valves and outputting in a single direction; s2, putting prawns, aquatic weeds and algae which form ecology with the prawns into each ecological unit, and arranging a water quality sensor and an ecological detector; s3, dynamically adjusting the water ecology; and S4, performing ecological treatment on the ecological units which do not meet the ecological balance. According to the invention, the culture area is divided into a plurality of ecological units, the ecological units are mutually adjusted to achieve dynamic ecological balance through a reinforcement learning algorithm, and then the ecological units which cannot achieve ecological balance are subjected to water ecological treatment, so that the drug and labor cost required by water ecological treatment can be greatly reduced, and the culture efficiency is improved.)

1. The ecological prawn breeding method is characterized by comprising the following steps:

s1, dividing the cultivation area into a plurality of ecological units, wherein the ecological units are separated by retaining walls and connected by pipelines controlled by control valves and outputting in a single direction;

s2, putting prawns, aquatic weeds and algae which form ecology with the prawns into each ecological unit, and arranging a water quality sensor and an ecological detector;

s3, detecting the environmental data of each ecological unit by a water quality sensor and an ecological detector, calculating the water ecological data and the data of the shrimp quantity which need to be adjusted when each ecological unit reaches ecological balance according to the current environmental data, making a decision by adopting a reinforcement learning algorithm, controlling a control valve to output a pipeline according to the decision method, outputting the water and the shrimps to be adjusted of the ecological unit to another ecological unit which needs to be adjusted, and dynamically adjusting the shrimp quantity and the water quality condition of each ecological unit;

and S4, performing ecological treatment on the ecological units which do not meet the ecological balance.

2. The ecological prawn breeding method according to claim 1, wherein in step S3, the reinforcement learning algorithm adopts a q-learning algorithm, the reinforcement learning algorithm includes individual agents and group agents, the individual agents process data into two tuples of state and strategy, the individual agents perform strategy adjustment of the prawn quantity and water quality of an ecological unit according to the state data of the current ecological unit and send the data to the group agents, and the group agents perform maximum mutual matching according to the data sent by each individual agent and control a control valve to perform pipeline output according to the decision method.

3. The ecological cultivation method for prawns according to claim 1, wherein in the step S2, the ecological monitor comprises a monitoring camera, the number of prawns, aquatic weeds, algae and protozoa is dynamically captured by using a feature extraction method, and the water quality sensor monitors the content of ammonia, nitrogen and oxygen in water and the ph value of water.

4. The ecological prawn breeding method according to claim 3, wherein the feature extraction is performed by using a convolutional neural network unit, and the convolutional neural network unit is composed of a network structure in the form of a convolutional unit which is normalized to a batch to activate a Relu function.

Technical Field

The invention relates to the technical field of prawn ecological breeding, in particular to a prawn ecological breeding method.

Background

The ecological breeding refers to a process of carrying out ecological cycle breeding by using aquatic organisms such as waterweeds and shrimps, wherein excrement generated by prawns can be absorbed by the waterweeds and the like, meanwhile, the waterweeds can generate nutrients for the prawns to form a cycle, but because the ecological environment has too many variable factors, the phenomenon that the ecological balance is easily broken in the process of carrying out ecological breeding is easily caused, the whole ecological environment can be damaged by inundation, the breeding efficiency of the prawns is influenced, therefore, the rebalance can be realized only by carrying out ecological treatment, but the existing ecological treatment has too many medicines, the required cost is higher, the treatment effect is poorer, and the ecological cycle breeding is not beneficial to the breeding of the prawns.

Disclosure of Invention

In order to solve at least or partially the problems, the method for ecologically culturing the prawns is provided, which can greatly reduce the cost of medicines and manpower required by water ecological treatment and improve the culture efficiency.

In order to achieve the purpose, the invention provides the following technical scheme:

the invention relates to an ecological prawn breeding method, which comprises the following steps:

s1, dividing the cultivation area into a plurality of ecological units, wherein the ecological units are separated by retaining walls and connected by pipelines controlled by control valves and outputting in a single direction;

s2, putting prawns, aquatic weeds and algae which form ecology with the prawns into each ecological unit, and arranging a water quality sensor and an ecological detector;

s3, detecting the environmental data of each ecological unit by a water quality sensor and an ecological detector, calculating the water ecological data and the data of the shrimp quantity which need to be adjusted when each ecological unit reaches ecological balance according to the current environmental data, making a decision by adopting a reinforcement learning algorithm, controlling a control valve to output a pipeline according to the decision method, outputting the water and the shrimps to be adjusted of the ecological unit to another ecological unit which needs to be adjusted, and dynamically adjusting the shrimp quantity and the water quality condition of each ecological unit;

and S4, performing ecological treatment on the ecological units which do not meet the ecological balance.

As a preferred technical solution of the present invention, in step S3, the reinforcement learning algorithm adopts a q-learning algorithm, the reinforcement learning algorithm includes individual agents and group agents, the individual agents process data into two tuples of state and policy, the individual agents perform policy adjustment of the shrimp number and water quality of the ecological unit according to the state data of the current ecological unit and send the data to the group agents, the group agents perform maximum mutual matching according to the data sent by each individual agent and control the control valve to perform pipeline output according to the decision method.

As a preferred embodiment of the present invention, in step S2, the ecology monitor includes a monitoring camera for dynamically capturing the number of shrimps, aquatic weeds, algae and protozoa by using a feature extraction method, and the water quality sensor monitors the content of ammonia, nitrogen and oxygen in water and the ph value of water.

As a preferred technical solution of the present invention, the feature extraction is performed by using a convolutional neural network unit, and the convolutional neural network unit is composed of a network structure in the form of a convolutional unit from convolution to batch normalization to activation of a Relu function.

Compared with the prior art, the invention has the following beneficial effects:

according to the invention, the culture area is divided into a plurality of ecological units, the ecological units are mutually adjusted to achieve dynamic ecological balance through a reinforcement learning algorithm, and then the ecological units which cannot achieve ecological balance are subjected to water ecological treatment, so that the drug and labor cost required by water ecological treatment can be greatly reduced, and the culture efficiency is improved.

Detailed Description

The following description of the preferred embodiments of the present invention is provided for the purpose of illustration and description, and is in no way intended to limit the invention.

In addition, if a detailed description of the known art is not necessary to show the features of the present invention, it is omitted.

Example 1

The invention provides an ecological prawn breeding method, which comprises the following steps:

s1, dividing the cultivation area into a plurality of ecological units, wherein the ecological units are separated by retaining walls and connected by pipelines controlled by control valves and outputting in a single direction;

s2, putting prawns, aquatic weeds and algae which form ecology with the prawns into each ecological unit, and arranging a water quality sensor and an ecological detector;

s3, detecting the environmental data of each ecological unit by a water quality sensor and an ecological detector, calculating the water ecological data and the data of the shrimp quantity which need to be adjusted when each ecological unit reaches ecological balance according to the current environmental data, making a decision by adopting a reinforcement learning algorithm, controlling a control valve to output a pipeline according to the decision method, outputting the water and the shrimps to be adjusted of the ecological unit to another ecological unit which needs to be adjusted, and dynamically adjusting the shrimp quantity and the water quality condition of each ecological unit;

and S4, performing ecological treatment on the ecological units which do not meet the ecological balance.

In the step S3, the reinforcement learning algorithm adopts a q learning algorithm, the reinforcement learning algorithm includes individual agents and group agents, the individual agents process data into two tuples of state and strategy, the individual agents perform strategy adjustment of the shrimp number and water quality condition of the ecological units according to the state data of the current ecological units and send the data to the group agents, the group agents perform maximum mutual matching according to the data sent by each individual agent and control the control valve to perform pipeline output according to the decision method.

In step S2, the ecological monitor includes a monitoring camera, the number of shrimps, aquatic weeds, algae and protozoa is dynamically captured by using a feature extraction method, and the water quality sensor monitors the content of ammonia, nitrogen and oxygen in water and the ph value of water.

The feature extraction is completed by adopting a convolution neural network unit, and the convolution neural network unit is composed of a network structure in a convolution unit form from convolution to batch standardization to activation of Relu function.

Specifically, each ecological unit is regarded as a closed ecological breeding area, because the ecological breeding areas can generate ecological change in the breeding process, for example, the water quality is poor due to excessive shrimps, the nutrients of aquatic plants and algae are less due to less shrimps, and the like, in the adjusting process, the water quality and the shrimp quantity of each ecological breeding area are adjusted, the ecological environment with more aquatic plants and algae is balanced due to poor water quality, so that a dynamic ecological balance is achieved, and after all other ecological breeding areas are balanced, the unbalanced ecological breeding areas are adjusted in an aquatic ecological treatment mode.

According to the invention, the culture area is divided into a plurality of ecological units, the ecological units are mutually adjusted to achieve dynamic ecological balance through a reinforcement learning algorithm, and then the ecological units which cannot achieve ecological balance are subjected to water ecological treatment, so that the drug and labor cost required by water ecological treatment can be greatly reduced, and the culture efficiency is improved.

Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.