Water ecology restoration plant community management method and system based on neural network

Abstract

The invention discloses a neural network-based water ecology restoration plant community management method and system, and the method comprises the steps: carrying out the water layering detection of a target water area based on an SOFM neural network, and obtaining a layering result; performing dynamic hydrological data sampling on each water body layer by utilizing water body monitoring equipment according to a layering result to obtain a sampling result of each water body layer; according to the aquatic plant distribution data and the aquatic animal distribution data, calling corresponding biological information and water body layer natural ecological environment data to construct a target database; based on a BP neural network, determining a dynamic change rule of the target water area within a preset time according to the target database; whether the dynamic change rule meets a preset restoration target or not is judged, according to the judgment result, based on an ecological restoration principle, the types and the corresponding number of plants needed by each water body layer are calculated, and a calculation result is obtained; and managing the plant community of the target water area by utilizing plant management equipment according to a calculation result.

Description

Technical Field

[0001] The present invention belongs to the technical field of water ecological restoration, and in particular relates to a water ecological restoration plant community management method and system based on a neural network. Background Art

[0002] One of the main reasons for the formation of black and smelly water bodies is the lack of oxygen in the water body, which is further exacerbated by water stratification. However, traditional water ecological restoration methods can only achieve horizontal circulation of surface water bodies and cannot improve the vertical stratification of water bodies, making it difficult to carry out deep-level water ecological restoration in deep water bodies.

[0003] In order to solve the problems existing in traditional water ecological restoration methods, the existing technology proposes a matching method based on water ecological restoration plant communities. After analyzing the natural data of the target water area, a suitable aquatic plant community is selected to manage the target water area. However, the water ecological restoration in the existing technology only considers the ecological restoration effect of aquatic plants on stratified water bodies, and does not consider the impact of aquatic animals in the target water area on aquatic plants, and the continuous impact of aquatic animals on the ecological environment of the target water area. Summary of the invention

[0004] In view of this, the purpose of the present invention is to provide a water ecological restoration plant community management method and system based on neural network to solve the above technical problems.

[0005] In order to achieve the above object, the present invention provides the following technical solutions: A water ecological restoration plant community management method based on neural network, comprising: Based on the SOFM neural network, the target water area is tested for water stratification and the stratification results are obtained; According to the stratification results, the water body monitoring equipment is used to sample the dynamic hydrological data of each water body layer to obtain the sampling results of each water body layer; the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water body layer; According to the distribution data of aquatic plants and aquatic animals, the corresponding biological information and the natural ecological environment data of the water layer are called to build the target database; Based on BP neural network, the dynamic change law of the target water area within the preset time is determined according to the target database; Determine whether the dynamic change law meets the preset restoration target. According to the judgment result and based on the principle of ecological restoration, calculate the plant species and corresponding quantity required for each water layer to obtain the calculation result; Based on the calculation results, plant management equipment is used to manage the plant community in the target water area.

[0006] Furthermore, the target water area is subjected to water stratification detection based on the SOFM neural network, and the stratification results are obtained, including: The controller is used to control the lifting motor to carry the water monitoring equipment to perform adaptive data collection in the target water area to obtain an initial natural data set; the initial natural data set is preprocessed to obtain a training data set; wherein the initial natural data includes dissolved oxygen data, temperature data and water pH value; Initialize the SOFM neural network, and iteratively train the initialized SOFM neural network according to the training data set until the number of iterative training meets the preset conditions to obtain the target SOFM neural network; Obtain the topological structure formed by the target SOFM neural network in the output layer, analyze the topological structure, realize the stratified detection of the target water area, and obtain the stratified results; Among them, the topological structure is used to reflect the relationship between the training data in the training data set; When analyzing the topological structure, each neuron group in the output layer of the target SOFM neural network represents a water layer of the target water area, and its corresponding weight vector represents the initial natural data of the current water layer.

[0007] Furthermore, according to the stratification results, the water body monitoring equipment is used to sample the dynamic hydrological data of each water body layer to obtain the sampling results of each water body layer, including: According to the stratification results, multiple water layers of the target waters are determined; Based on the random sampling method, multiple sampling areas are set for each water layer of the target water area; Use water monitoring equipment to sample dynamic hydrological data in each sampling area and obtain multiple dynamic hydrological sampling results for each water layer; Integrate multiple dynamic hydrological sampling results of each water layer to obtain a unique sampling result for each water layer; the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of each water layer; Aquatic plant distribution data include aquatic plant category distribution data and aquatic plant quantity distribution data; Aquatic animal distribution data include aquatic animal category distribution data and aquatic animal quantity distribution data; The natural ecological environment data of the water layer include dissolved oxygen data, temperature data and water pH value.

[0008] Furthermore, according to the aquatic plant distribution data and aquatic animal distribution data, the corresponding biological information and the natural ecological environment data of the water layer are called to construct a target database, including: Obtain aquatic plant category data from aquatic plant distribution data; Obtain aquatic animal category data from aquatic animal distribution data; According to the aquatic plant category data and aquatic animal category data, based on network communication methods, the corresponding biological information and water layer natural ecological environment data are called to build a target database; Among them, the corresponding biological information includes aquatic plant information and aquatic animal information; Aquatic plant information includes plant type, suitable water depth, suitable water temperature, suitable water pH value and dissolved oxygen growth rate; Aquatic animal information includes animal type, suitable water depth, suitable water temperature, suitable water pH value, oxygen consumption rate, gnawing plant species and plant gnawing efficiency.

[0009] Furthermore, based on the BP neural network, the dynamic change law of the target water area within a preset time is determined according to the target database, including: Obtain historical dynamic hydrological data information in several waters and construct a training data set; the dynamic hydrological data information includes aquatic plant data, aquatic animal data and natural ecological environment data of each water layer in several waters at different time scales; Aquatic plant data include aquatic plant distribution data and aquatic plant information; Aquatic animal data include aquatic animal distribution data and aquatic animal information; Based on the training data set, the BP neural network is used to simulate the response of aquatic plants and aquatic animals in different water layers to the changes in the natural ecological environment data of each water layer at different time scales without human interference, and a BP neural network prediction model is generated; The information in the target database is obtained as input data and input into the BP neural network prediction model, and the dynamic change rules among the aquatic plant data, aquatic animal data and the natural ecological environment data of the target water area are obtained as output.

[0010] Further, judging whether the dynamic change law meets the preset repair target includes: Obtain the dynamic change rules among aquatic plant data, aquatic animal data and natural ecological environment data in the target waters; Obtain the target time scale, and determine the target time node in the dynamic change law according to the target time scale; Obtain the target natural ecological environment data corresponding to the dynamic change law according to the target time node; Determine whether the target natural ecological environment meets the preset restoration goals and generate a judgment result; Among them, the judgment result is that the target natural ecological environment meets the preset restoration target; Or, the target natural ecological environment does not meet the preset restoration goals.

[0011] Furthermore, according to the judgment results and based on the principle of ecological restoration, the plant species and corresponding quantities required for each water layer are calculated to obtain the calculation results, including: Get the judgment result. If the judgment result is that the target natural ecological environment does not meet the preset restoration target, get the ecological restoration principles: in, is the preset restoration target of the j-th water layer, is the number of aquatic plant species in the jth water layer, is the matching coefficient of the i-th aquatic plant in the j-th water layer, is the total dissolved oxygen growth rate of the i-th aquatic plant in the target waters per unit time, is the average growth efficiency of the i-th aquatic plant per unit time, is the average grazing efficiency of the kth aquatic animal on the ith aquatic plant per unit time, is the total oxygen consumption rate of the kth aquatic animal in the target waters per unit time; when When , it means that the matching relationship between the i-th aquatic plant and the j-th water layer is a match. When , it means that the matching relationship between the i-th aquatic plant and the j-th water layer is mismatched. Indicates that the suitable water depth, suitable water temperature, and suitable water pH value of the i-th aquatic plant match the temperature data and water pH value of the j-th water layer; Calculate the plant species and corresponding quantities required for each water layer so that the left and right equations of the ecological restoration principle are equal and obtain the calculation results.

[0012] Furthermore, according to the calculation results, plant management equipment is used to manage the plant community in the target waters, including: Obtain the required plant species and corresponding quantities for each water layer included in the calculation results; According to the difference of water layer, call the corresponding plant management equipment; Control the corresponding plant management equipment to load the plant species and corresponding quantity required for the corresponding water layer, and go to the corresponding position of the target water area to arrange the plants; In the process of obtaining the plant species and corresponding quantities required for calculating each water layer, Corresponding to the distribution position of the i-th aquatic plant, the corresponding aquatic plant harvesting system is called to cut and recycle the i-th aquatic plant.

[0013] Furthermore, a neural network-based water ecological restoration plant community management also includes: The moment when the plant management equipment manages the plant community in the target water area and the management ends is taken as the starting time point; Obtain historical dissolved oxygen data and historical biological content information of different water layers in the target water area after a preset second time from the starting time point; Analyze historical dissolved oxygen data to obtain the historical dissolved oxygen change rate of different water layers per unit time; Analyze the historical biomass content information to obtain the historical biomass content information of each organism in different water layers per unit time; Obtain the dissolved oxygen change rate and biological content information of different water layers after a unit time at the current moment; Obtain dynamic change rules and extract standard biological content information of different water layers after a unit time at the current moment; The water ecological restoration detection score of the current target water area is calculated based on the historical dissolved oxygen change rate, historical biological content information, dissolved oxygen change rate, biological content information and standard biological content information. The water ecological restoration detection score and the preset threshold are used to determine whether the current water ecological restoration progress meets the preset restoration progress.

[0014] A water ecological restoration plant community management system based on neural network, comprising: The water body stratification detection module is used to perform water body stratification detection on the target water area according to the SOFM neural network to obtain stratification results; The hydrological data sampling module is used to perform dynamic hydrological data sampling on each water layer using water monitoring equipment according to the stratification results, and obtain the sampling results of each water layer; wherein the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water layer; The database module is used to construct a target database based on the aquatic plant distribution data and the aquatic animal distribution data, calling the corresponding biological information and the natural ecological environment data of the water layer; The water area data dynamic change monitoring module is used to determine the dynamic change law of the target water area within a preset time according to the target database through the BP neural network; The plant community restoration calculation module is used to determine whether the dynamic change law meets the preset restoration target. According to the judgment result and based on the principle of ecological restoration, the plant species and corresponding quantity required for each water layer are calculated to obtain the calculation result; The plant community management module is used to manage the plant community in the target water area using plant management equipment according to the calculation results.

[0015] The beneficial effects of the present invention are: The present invention proposes a water ecological restoration plant community management method based on a neural network, which is used to accurately stratify the target water area according to the pollution conditions of different water body layers by using a SOFM neural network, which is beneficial to improving the efficiency of subsequent water ecological restoration. At the same time, in the process of water ecological restoration, a BP neural network is used to predict the dynamic change law of the dependency relationship between animals, plants and natural data, and dynamic restoration of the water ecological environment is achieved through the dynamic change law, which is beneficial to improving the reliability of ecological environmental restoration in the target water area, and finally realizes the completion of ecological self-circulation restoration for different water body layers, thereby achieving the purpose of water ecological restoration.

[0016] Other advantages, objectives and features of the present invention will be described in the following description and will be apparent to those skilled in the art to some extent, or they may be taught from the practice of the present invention. The purposes and other advantages of the present invention may be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0017] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings: Figure 1 A method flow chart of a water ecological restoration plant community management method and system based on a neural network in an embodiment of the present invention; Figure 2 A flow chart of a water ecological restoration plant community management method based on a neural network and a sampling process of a water layer in the system in an embodiment of the present invention; Figure 3 Schematic diagram of system modules of a neural network-based water ecological restoration plant community management method and system in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0020] See also Figure 1 The present invention proposes a water ecological restoration plant community management method based on a neural network, comprising: S101, performing water body stratification detection on the target water area based on the SOFM neural network to obtain stratification results; S102. According to the stratification results, use water body monitoring equipment to sample dynamic hydrological data for each water body layer to obtain sampling results for each water body layer; wherein the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water body layer; S103, based on the aquatic plant distribution data and the aquatic animal distribution data, call the corresponding biological information and the natural ecological environment data of the water body layer to build a target database; S104, based on the BP neural network, determining the dynamic change law of the target water area within a preset time according to the target database; S105, judging whether the dynamic change law meets the preset restoration target, and according to the judgment result and based on the principle of ecological restoration, calculating the plant species and corresponding quantity required for each water layer to obtain the calculation result; S106. Using plant management equipment to manage the plant community in the target water area according to the calculation results; The working principle of the above technical solution is: in the actual water ecological environment, when an area can complete a perfect ecological self-circulation, it must satisfy the ecological circulation system of animals and plants. However, the water ecological restoration in the prior art only uses plants to perform ecological restoration of water bodies in layers, and does not combine the actual water ecological environment for dynamic restoration. Therefore, the present invention proposes a water ecological restoration plant community management method based on a neural network to solve the above problem; Firstly, the SOFM neural network is used to cluster the hydrological data, and the target water area is tested for water stratification, and the area with obvious eco-hydrological characteristics is divided to obtain the stratification test results. The existing water stratification includes the surface layer, thermocline layer, isothermal layer, bottom layer and sedimentary layer. Compared with the existing conventional water stratification method, this method can perform cluster analysis according to the eco-hydrological characteristics of the polluted water body, and accurately stratify the water body according to the specific pollution data, so as to facilitate the subsequent arrangement of the corresponding number and type of plants to accurately repair the water stratification. The polluted water body mainly described in this paper is the black and smelly water body formed due to lack of oxygen in the water body. After the test results are obtained, the water body monitoring equipment is used to sample the dynamic hydrological data of each water body layer according to the test results to obtain the sampling results of each water body layer, that is, the dynamic hydrological data of each water body layer; wherein the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water body layer; The aquatic plant distribution data and aquatic animal distribution data are the distribution density of different plants and animals in each water layer and the distribution position of the water layer. It is worth noting that each water layer is preferably partitioned according to the type of plant or animal, and then the distribution density of the corresponding plant or animal is calculated according to the partition result, and the distribution location is determined at the same time; The natural ecological environment data of the water layer include but are not limited to the dissolved oxygen data, temperature data, water flow data, water pH value and other natural data; Then, based on the aquatic plant distribution data and aquatic animal distribution data, the basic data of each individual in the corresponding biological information and the natural ecological environment data of all water layers are called by the network or other feasible links to construct a target database; this step is beneficial for individual monitoring of all groups in the target water area, and facilitates the subsequent controllable management of the plant community in the target water area; Then, the BP neural network is used to determine the dynamic change law of the target water area within the preset time according to the target database, that is, to predict the change law of oxygen content in the target water area in the subsequent time period without human interference; And judge whether the dynamic change law meets the preset restoration target. According to the judgment result and based on the principle of ecological restoration, calculate the plant species and corresponding quantity required for each water layer to obtain the calculation result; Based on the calculation results, plant management equipment is used to manage the plant community in the target waters; management includes but is not limited to various methods such as harvesting, planting and transfer; The beneficial effects of the above technical scheme are as follows: through the above technical scheme, the SOFM neural network is used to accurately stratify the target waters according to the pollution conditions of different water layers, which is beneficial to improving the efficiency of subsequent water ecological restoration. At the same time, in the process of water ecological restoration, the BP neural network is used to predict the dynamic change law of the dependency relationship between animals, plants and natural data, and the dynamic restoration of the water ecological environment is achieved through the dynamic change law, which is beneficial to improving the reliability of the ecological environment restoration of the target waters, and finally achieving the completion of ecological self-circulation restoration for different water layers, thereby achieving the purpose of water ecological restoration.

[0021] In one embodiment, the target water area is subjected to water body stratification detection based on the SOFM neural network to obtain stratification results, including: The controller is used to control the lifting motor to carry the water monitoring equipment to perform adaptive data collection in the target waters to obtain an initial natural data set; the initial natural data set is preprocessed to obtain a training data set; wherein the initial natural data includes dissolved oxygen data, temperature data and water pH value; Initialize the SOFM neural network, and iteratively train the initialized SOFM neural network according to the training data set until the number of iterative training meets the preset conditions to obtain the target SOFM neural network; Obtain the topological structure formed by the target SOFM neural network in the output layer, analyze the topological structure, realize the stratified detection of the target water area, and obtain the stratified results; Among them, the topological structure is used to reflect the relationship between the training data in the training data set; When analyzing the topological structure, each neuron group in the output layer of the target SOFM neural network represents a water layer of the target water area, and its corresponding weight vector represents the initial natural data of the current water layer; The working principle of the above technical solution is as follows: before carrying out water ecological restoration in the target water area, first use the controller to control the lifting motor to carry the water monitoring equipment to perform adaptive data collection in the target water area to obtain an initial natural data set; pre-process the initial natural data set to obtain a training data set; wherein the initial natural data includes but is not limited to dissolved oxygen data, temperature data and water pH value, and according to specific ecological restoration standards, also includes flow rate, water turbidity and other measurement data; the collected data needs to be pre-processed to ensure that it is suitable for input into the SOFM network, and the pre-processing includes but is not limited to data cleaning, normalization, and removal of outliers; initialize the SOFM neural network, in which each neuron or node is associated with a weight vector, initialize the SOFM neural network so that these weight vectors are randomly assigned in the initial stage; then iteratively train the initialized SOFM neural network according to the training data set until the number of iterative training meets the preset conditions and the target is obtained. SOFM neural network; during the iteration process, the SOFM neural network will learn the intrinsic structure of the training data set. In each iteration, the SOFM neural network compares the initial natural data with the weight vector of each neuron, finds the neuron that best matches it, and adjusts its weight to better match the input initial natural data. At the same time, the neurons around the successfully matched neurons will also adjust their weights accordingly. When the number of iterative training meets the preset conditions, the SOFM neural network can map similar initial natural data to adjacent neurons, thereby forming a topological structure in the output layer. At this time, the topological structure formed by the target SOFM neural network in the output layer is obtained, and the topological structure is analyzed to realize the stratification detection of the target water area and obtain the stratification result. The topological structure is used to reflect the relationship between the training data in the training data set; when analyzing the topological structure, each neuron group in the output layer of the target SOFM neural network represents a water layer of the target water area, and its corresponding weight vector represents the initial natural data of the current water layer; The beneficial effect of the above technical scheme is: through the above technical scheme, the SOFM neural network is used to accurately stratify the target waters according to the pollution conditions of different water layers, which is beneficial to improving the efficiency of subsequent water ecological restoration and achieving efficient management of different water layers in the target waters.

[0022] See also Figure 2 In one embodiment, according to the stratification results, the water body monitoring equipment is used to perform dynamic hydrological data sampling on each water body layer to obtain the sampling results of each water body layer, including: S201, determining multiple water layers of the target water area according to the stratification results; It is worth noting that when determining the water layer, it is also necessary to determine the water depth range of each water layer to facilitate the subsequent management of aquatic plants; S202, based on a random sampling method, setting multiple sampling areas for each water layer of the target water area; When random sampling is performed, the number and range of sampling areas depends on the size of the target water area. The larger the target water area, the larger the number and range of sampling areas. S203, using water body monitoring equipment to sample dynamic hydrological data in each sampling area, and obtaining multiple dynamic hydrological sampling results for each water body layer; Dynamic hydrological data sampling includes collecting information on aquatic plants, aquatic animals and natural ecological environment; S204, integrating multiple dynamic hydrological sampling results of each water layer to obtain a unique sampling result for each water layer; wherein the sampling result includes aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of each water layer; The integration operation is preferably to integrate the types in the aquatic plant distribution data and the aquatic animal distribution data without duplication, to statistically integrate their distribution data, and to calculate and integrate the average value of the natural ecological environment data of the water body layer; Aquatic plant distribution data include aquatic plant category distribution data and aquatic plant quantity distribution data; Aquatic animal distribution data include aquatic animal category distribution data and aquatic animal quantity distribution data; The natural ecological environment data of the water layer include dissolved oxygen data, temperature data and water pH value; The beneficial effect of the above technical solution is: through the above technical solution, the biological data and natural data of each water layer are collected, which is beneficial to improving the reliability of subsequent water ecological restoration.

[0023] In one embodiment, according to the aquatic plant distribution data and the aquatic animal distribution data, the corresponding biological information and the natural ecological environment data of the water body layer are called to construct a target database, including: Obtain aquatic plant category data from aquatic plant distribution data; Obtain aquatic animal category data from aquatic animal distribution data; According to the aquatic plant category data and aquatic animal category data, based on network communication methods, the corresponding biological information and water layer natural ecological environment data are called to build a target database; The target database includes the natural ecological environment data, aquatic plant distribution data, aquatic animal distribution data and corresponding biological information corresponding to each water layer; Among them, the corresponding biological information includes aquatic plant information and aquatic animal information; Aquatic plant information includes but is not limited to plant type, suitable water depth, suitable water temperature, suitable water pH value and dissolved oxygen growth rate; The dissolved oxygen growth rate is obtained by calling the corresponding data in the existing target database, or by detecting the dissolved oxygen concentration at different periods by an instrument and calculating and storing it in the target database, and then calling it from the target database; Aquatic animal information includes but is not limited to animal type, suitable water depth, suitable water temperature, suitable water pH value, oxygen consumption rate, gnawing plant species and plant gnawing efficiency; The beneficial effects of the above technical solution are: through the above technical solution, a corresponding target ecological database is established for the target water area, which improves the comprehensive understanding of the target water area ecosystem and provides reliable data support for subsequent water ecological governance and water area management.

[0024] In one embodiment, based on the BP neural network, the dynamic change law of the target water area within a preset time is determined according to the target database, including: Obtain historical dynamic hydrological data information in several waters and construct a training data set; the dynamic hydrological data information includes aquatic plant data, aquatic animal data and natural ecological environment data of each water layer in several waters at different time scales; Aquatic plant data include aquatic plant distribution data and aquatic plant information; The distribution data of aquatic plants include the water layer in which they are distributed and the corresponding quantity; Aquatic animal data include aquatic animal distribution data and aquatic animal information; The distribution data of aquatic animals include the water layers in which they are distributed and the corresponding numbers; Based on the training data set, the BP neural network is used to simulate the response of aquatic plants and aquatic animals in different water layers to the changes in the natural ecological environment data of each water layer at different time scales without human interference, and a BP neural network prediction model is generated; The information in the target database is obtained as input data, which is input into the BP neural network prediction model, and the dynamic change rules among the aquatic plant data, aquatic animal data and the natural ecological environment data in the target water area are obtained as output; The working principle of the above technical solution is as follows: obtaining dynamic hydrological data information of different water layers of several water bodies without human interference, including aquatic plant data, aquatic animal data and natural ecological environment data of each water layer of several water bodies at different time scales, such as information on plant and animal species, quantity, water quality parameters of the water body, etc., to construct a training data set; preprocessing the collected training data set is beneficial to improving the training efficiency and prediction accuracy of the model; then designing the structure of the BP neural network, and using the BP neural network to simulate the response rules of aquatic plants and aquatic animals in different water layers at different time scales to changes in the natural ecological environment data of each water layer without human interference according to the training data set, and generate a BP neural network prediction model. Preferably, the BP neural network The number of nodes in the input layer of the network structure matches the characteristic number of the corresponding aquatic plant data, aquatic animal data and natural ecological environment data of each water layer in the water area, and the number of nodes in the output layer matches the corresponding response law. The specific simulation method preferably refers to the application of BP neural network in the delayed response of desert rodent communities to climate change based on BP neural network, which will not be elaborated here. Then, the information in the target database is obtained as input data, which is input into the BP neural network prediction model, and the dynamic change law among the aquatic plant data, aquatic animal data and natural ecological environment data in the target water area is output. It is worth mentioning that a curve chart can also be generated according to the dynamic change law, which is beneficial for managers to intuitively feel the change trend among the three. Furthermore, in addition to using the BP neural network to determine the dynamic change law among the aquatic plant data, aquatic animal data and the natural ecological environment data of the water area in the target water area, the oxygenation efficiency of the aquatic plant data of each layer in the target water area, the oxygen consumption efficiency of the aquatic animal and the original oxygen content in the natural ecological environment data of the water area can also be counted to generate a curve change diagram of the water area oxygen content per unit time, which is used to determine the dynamic change law among the aquatic plant data, aquatic animal data and the natural ecological environment data of the water area in the target water area; The beneficial effects of the above technical scheme are: through the above technical scheme, the BP neural network is used to predict the dynamic change law of the dependency relationship between animals, plants and natural data, and the dynamic restoration of the water ecological environment is achieved through the dynamic change law, which is beneficial to improving the reliability of the ecological environment restoration of the target water area.

[0025] In one embodiment, determining whether the dynamic change rule meets the preset repair target includes: Obtain the dynamic change rules among aquatic plant data, aquatic animal data and natural ecological environment data in the target waters; Obtain the target time scale, and determine the target time node in the dynamic change law according to the target time scale; The target time scale is a time period set by the user. The target time node in the dynamic change law is determined according to the target time scale, that is, the last time node of the time period is determined as the target time node; According to the target time node, the target natural ecological environment data corresponding to the dynamic change law is obtained, that is, the natural ecological environment data corresponding to the target time node in the curve graph generated by the dynamic change law is obtained as the target natural ecological environment data; Determine whether the target natural ecological environment meets the preset restoration goals and generate a judgment result; The preset restoration target is preferably the oxygen content threshold of each water layer in the water area; Among them, the judgment result is that the target natural ecological environment meets the preset restoration target; Or, the target natural ecological environment does not meet the preset restoration goals.

[0026] In one embodiment, according to the judgment result and based on the principle of ecological restoration, the plant species and corresponding quantity required for each water layer are calculated to obtain the calculation results, including: Get the judgment result. If the judgment result is that the target natural ecological environment does not meet the preset restoration target, get the ecological restoration principles: in, is the preset restoration target of the j-th water layer, is the number of aquatic plant species in the jth water layer, is the matching coefficient of the i-th aquatic plant in the j-th water layer, is the total dissolved oxygen growth rate of the i-th aquatic plant in the target waters per unit time, is the average growth efficiency of the i-th aquatic plant per unit time, is the average grazing efficiency of the kth aquatic animal on the ith aquatic plant per unit time, is the total oxygen consumption rate of the kth aquatic animal in the target waters per unit time; when When , it means that the matching relationship between the i-th aquatic plant and the j-th water layer is a match. When , it means that the matching relationship between the i-th aquatic plant and the j-th water layer is mismatched. Indicates that the suitable water depth, suitable water temperature, and suitable water pH value of the i-th aquatic plant match the temperature data and water pH value of the j-th water layer; the suitable water depth matches the corresponding water layer; Calculate the plant species and corresponding quantity required for each water layer, so that the left and right equations of the ecological restoration principle are equal, or the result on the left is greater than the right, and calculate the corresponding number and species of plants that need to be added. It is worth noting that there is no limit on the species and quantity of plants that need to be added, as long as they can meet the preset restoration goals; The beneficial effect of the above technical solution is: through the above technical solution, reliable data support is provided for plant management in the target waters.

[0027] In one embodiment, according to the calculation results, the plant community in the target water area is managed using the plant management device, including: Obtain the required plant species and corresponding quantities for each water layer included in the calculation results; According to the difference of water layer, call the corresponding plant management equipment; Since aquatic plants are usually divided into floating plants, emergent plants and submerged plants, different aquatic plants need to be planted for different water layers. Therefore, it is necessary to call corresponding plant management equipment to plant and manage aquatic plants according to the difference in water layers; Control the corresponding plant management equipment to load the plant species and corresponding quantity required for the corresponding water layer, and go to the corresponding position of the target water area to arrange the plants; At the same time, in the process of obtaining the plant species and corresponding quantities required for calculating each water layer, The corresponding distribution position of the i-th aquatic plant is called to cut and recycle the i-th aquatic plant by the corresponding aquatic plant harvesting system; Since the i-th aquatic plant is not suitable for growing in the corresponding water layer, its growth trend in the water layer is relatively slow, and it will also affect other aquatic plants in the water layer, such as blocking sunlight, which in turn affects the ecological restoration efficiency of the water layer. Therefore, choosing to cut and recycle it is beneficial to improve the standardization of plant management; Furthermore, aquatic plants can be regulated and managed so that they grow in matching water layers; The beneficial effect of the above technical solution is: through the above technical solution, the management of plant communities is achieved.

[0028] In one embodiment, a neural network-based water ecological restoration plant community management further includes: The moment when the plant management equipment manages the plant community in the target water area and the management ends is taken as the starting time point; Obtain historical dissolved oxygen data and historical biological content information of different water layers in the target water area after a preset second time from the starting time point; Analyze historical dissolved oxygen data to obtain the historical dissolved oxygen change rate of different water layers per unit time; Analyze the historical biomass content information to obtain the historical biomass content information of each organism in different water layers per unit time; After obtaining the unit time at the current moment, the dissolved oxygen change rate of different water layers and the biological content information of different water layers in the current detection period are obtained; Obtain dynamic change rules and extract standard biological content information of different water layers after a unit time at the current moment; The standard biomass content information is preferably determined based on a curve chart generated according to a dynamic change rule, or directly based on the preset standard biomass content information of the management personnel; The water ecological restoration test score of the current target water area is calculated based on the historical dissolved oxygen change rate, historical biological content information, dissolved oxygen change rate, biological content information and standard biological content information. The water ecological restoration test score and the preset threshold are used to determine whether the current water ecological restoration progress meets the preset restoration progress; The calculation formula is preferably: Among them, F represents the water ecological restoration detection score of the current target water area, It represents the water ecological restoration detection score of the j-th water layer in the current target water area; and is the preset weight, Indicates the dissolved oxygen detection score of water ecological restoration, It indicates the score of water ecological restoration biological information detection; represents the dissolved oxygen detection score of the water ecological restoration of the j-th water layer, represents the water ecological restoration biological information detection score of the j-th water layer; Divide the unit time into multiple moments according to the preset moment standard. Indicates the number of moments that pass when the current moment passes the unit time. The number of moments is determined according to the length of the unit time. When the length of the unit time is determined, the number of moments is fixed; represents the number of moments when the jth water layer is detected; Each detection within a unit time is regarded as a detection cycle. Indicates the number of The dissolved oxygen change rate at each moment, represents the number of historical detection cycles, Indicates the number of the cth historical detection cycle The historical dissolved oxygen change rate at each moment; Indicates the jth water layer in the current detection cycle The dissolved oxygen change rate at each moment, represents the number of historical detection cycles of the jth water layer. Usually, the number of historical detection cycles of all water layers is the same. Indicates that the jth water layer is the first in the cth historical detection cycle The historical dissolved oxygen change rate at each moment; Indicates the number of species of aquatic plants and animals in the target waters. Indicates the number of The biological content information of the bth organism in the detection information at the moment, Indicates the standard biological content information of the bth organism, Indicates the number of the cth historical detection cycle The historical biological content information of the bth organism in the detection information at the moment; represents the number of aquatic plants and animals in the jth water layer in the target water area, Indicates the jth water layer in the current detection cycle The biological content information of the bth organism in the detection information at the moment, represents the standard biomass content information of the bth organism in the jth water layer, represents the jth water layer in the cth historical detection cycle The historical biological content information of the bth organism in the detection information at the moment; If the water ecological restoration detection score is less than the preset threshold, it is determined that the current water ecological restoration progress does not meet the preset restoration progress, and an alarm signal is fed back to the management personnel to facilitate the management personnel to survey the target waters; The beneficial effect of the above technical scheme is: through the above technical scheme, the ecological restoration progress of the target water area can be further monitored automatically. Compared with the existing technology of treating the target water area once and then letting it develop, this scheme is beneficial to improving the reliability of water ecological restoration.

[0029] See also Figure 3 , a water ecological restoration plant community management system based on neural network, comprising: The water body stratification detection module is used to perform water body stratification detection on the target water area according to the SOFM neural network to obtain stratification results; The hydrological data sampling module is used to perform dynamic hydrological data sampling on each water layer using water monitoring equipment according to the stratification results, and obtain the sampling results of each water layer; wherein the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water layer; The database module is used to construct a target database based on the aquatic plant distribution data and the aquatic animal distribution data, calling the corresponding biological information and the natural ecological environment data of the water layer; The water area data dynamic change monitoring module is used to determine the dynamic change law of the target water area within a preset time according to the target database through the BP neural network; The plant community restoration calculation module is used to determine whether the dynamic change law meets the preset restoration target. According to the judgment result and based on the principle of ecological restoration, the plant species and corresponding quantity required for each water layer are calculated to obtain the calculation result; The plant community management module is used to manage the plant community in the target water area using plant management equipment according to the calculation results.

[0030] The working principle and beneficial effects of the above technical solution have been explained in the method claims and will not be repeated here.

[0031] Finally, it should be noted that the above preferred embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail through the above preferred embodiments, those skilled in the art should understand that various changes can be made in form and details without departing from the scope defined by the claims of the present invention.

Claims

1. A method for managing plant communities for water ecological restoration based on neural networks, characterized in that: include: Based on the SOFM neural network, the target water area is tested for water stratification and the stratification results are obtained; According to the stratification results, the water body monitoring equipment is used to sample the dynamic hydrological data of each water body layer to obtain the sampling results of each water body layer; the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water body layer; According to the distribution data of aquatic plants and aquatic animals, the corresponding biological information and the natural ecological environment data of the water layer are called to build the target database; Based on BP neural network, the dynamic change law of the target water area within the preset time is determined according to the target database; Determine whether the dynamic change law meets the preset restoration target. According to the judgment result and based on the principle of ecological restoration, calculate the plant species and corresponding quantity required for each water layer to obtain the calculation result; Based on the calculation results, plant management equipment is used to manage the plant community in the target water area.

2. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: Based on the SOFM neural network, the target water area is tested for water stratification and the stratification results are obtained, including: The controller is used to control the lifting motor to carry the water monitoring equipment to perform adaptive data collection in the target water area to obtain an initial natural data set; the initial natural data set is preprocessed to obtain a training data set; wherein the initial natural data includes water body dissolved oxygen data, temperature data and pH value; Initialize the SOFM neural network, and iteratively train the initialized SOFM neural network according to the training data set until the number of iterative training meets the preset conditions to obtain the target SOFM neural network; Obtain the topological structure formed by the target SOFM neural network in the output layer, analyze the topological structure, realize the stratified detection of the target water area, and obtain the stratified results; Among them, the topological structure is used to reflect the relationship between the training data in the training data set; When analyzing the topological structure, each neuron group in the output layer of the target SOFM neural network represents a water layer of the target water area, and its corresponding weight vector represents the initial natural data of the current water layer.

3. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: According to the stratification results, the water monitoring equipment is used to sample the dynamic hydrological data of each water layer to obtain the sampling results of each water layer, including: According to the stratification results, multiple water layers of the target waters are determined; Based on the random sampling method, multiple sampling areas are set for each water layer of the target water area; Use water monitoring equipment to sample dynamic hydrological data in each sampling area and obtain multiple dynamic hydrological sampling results for each water layer; Integrate multiple dynamic hydrological sampling results of each water layer to obtain a unique sampling result for each water layer; the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of each water layer; Aquatic plant distribution data include aquatic plant category distribution data and aquatic plant quantity distribution data; Aquatic animal distribution data include aquatic animal category distribution data and aquatic animal quantity distribution data; The natural ecological environment data of the water layer include dissolved oxygen data, temperature data and water pH value.

4. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: According to the distribution data of aquatic plants and aquatic animals, the corresponding biological information and the natural ecological environment data of the water layer are called to build the target database, including: Obtain aquatic plant category data from aquatic plant distribution data; Obtain aquatic animal category data from aquatic animal distribution data; According to the aquatic plant category data and aquatic animal category data, based on network communication methods, the corresponding biological information and water layer natural ecological environment data are called to build a target database; Among them, the corresponding biological information includes aquatic plant information and aquatic animal information; Aquatic plant information includes plant type, suitable water depth, suitable water temperature, suitable water pH value and dissolved oxygen growth rate; Aquatic animal information includes animal type, suitable water depth, suitable water temperature, suitable water pH value, oxygen consumption rate, gnawing plant species and plant gnawing efficiency.

5. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: Based on BP neural network, the dynamic change law of the target water area within the preset time is determined according to the target database, including: Obtain historical dynamic hydrological data information in several waters and construct a training data set; the dynamic hydrological data information includes aquatic plant data, aquatic animal data and natural ecological environment data of each water layer in several waters at different time scales; Aquatic plant data include aquatic plant distribution data and aquatic plant information; Aquatic animal data include aquatic animal distribution data and aquatic animal information; Based on the training data set, the BP neural network is used to simulate the response of aquatic plants and aquatic animals in different water layers to the changes in the natural ecological environment data of each water layer at different time scales without human interference, and a BP neural network prediction model is generated; The information in the target database is obtained as input data and input into the BP neural network prediction model, and the dynamic change rules among the aquatic plant data, aquatic animal data and the natural ecological environment data of the target water area are obtained as output.

6. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: Determine whether the dynamic change rules meet the preset repair goals, including: Obtain the dynamic change rules among aquatic plant data, aquatic animal data and natural ecological environment data in the target waters; Obtain the target time scale, and determine the target time node in the dynamic change law according to the target time scale; Obtain the target natural ecological environment data corresponding to the dynamic change law according to the target time node; Determine whether the target natural ecological environment meets the preset restoration goals and generate a judgment result; Among them, the judgment result is that the target natural ecological environment meets the preset restoration target; Or, the target natural ecological environment does not meet the preset restoration goals.

7. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: According to the judgment results and based on the principle of ecological restoration, the plant species and corresponding quantities required for each water layer are calculated to obtain the calculation results, including: Get the judgment result. If the judgment result is that the target natural ecological environment does not meet the preset restoration target, get the ecological restoration principles: in, is the preset restoration target of the j-th water layer, is the number of aquatic plant species in the jth water layer, is the matching coefficient of the i-th aquatic plant in the j-th water layer, is the total dissolved oxygen growth rate of the i-th aquatic plant in the target waters per unit time, is the average growth efficiency of the i-th aquatic plant per unit time, is the average grazing efficiency of the kth aquatic animal on the ith aquatic plant per unit time, is the total oxygen consumption rate of the kth aquatic animal in the target waters per unit time; when When , it means that the matching relationship between the i-th aquatic plant and the j-th water layer is a match. When , it means that the matching relationship between the i-th aquatic plant and the j-th water layer is mismatched. Indicates that the suitable water depth, suitable water temperature, and suitable water pH value of the i-th aquatic plant match the temperature data and water pH value of the j-th water layer; Calculate the plant species and corresponding quantities required for each water layer so that the left and right equations of the ecological restoration principle are equal and obtain the calculation results.

8. The method for managing water ecological restoration plant communities based on neural networks according to claim 1 is characterized in that: Based on the calculation results, plant management equipment is used to manage the plant community in the target waters, including: Obtain the required plant species and corresponding quantities for each water layer included in the calculation results; According to the difference of water layer, call the corresponding plant management equipment; Control the corresponding plant management equipment to load the plant species and corresponding quantity required for the corresponding water layer, and go to the corresponding position of the target water area to arrange the plants; In the process of obtaining the plant species and corresponding quantities required for calculating each water layer, Corresponding to the distribution position of the i-th aquatic plant, the corresponding aquatic plant harvesting system is called to cut and recycle the i-th aquatic plant.

9. The method for managing water ecological restoration plant communities based on neural networks according to claim 1, characterized in that: Also includes: The moment when the plant management equipment manages the plant community in the target water area and the management ends is taken as the starting time point; Obtain historical dissolved oxygen data and historical biological content information of different water layers in the target water area after a preset second time from the starting time point; Analyze historical dissolved oxygen data to obtain the historical dissolved oxygen change rate of different water layers per unit time; Analyze the historical biomass content information to obtain the historical biomass content information of each organism in different water layers per unit time; Obtain the dissolved oxygen change rate and biological content information of different water layers after a unit time at the current moment; Obtain dynamic change rules and extract standard biological content information of different water layers after a unit time at the current moment; The water ecological restoration detection score of the current target water area is calculated based on the historical dissolved oxygen change rate, historical biological content information, dissolved oxygen change rate, biological content information and standard biological content information. The water ecological restoration detection score and the preset threshold are used to determine whether the current water ecological restoration progress meets the preset restoration progress.

10. A water ecological restoration plant community management system based on neural network, characterized in that: include: The water body stratification detection module is used to perform water body stratification detection on the target water area according to the SOFM neural network to obtain stratification results; The hydrological data sampling module is used to perform dynamic hydrological data sampling on each water layer using water monitoring equipment according to the stratification results, and obtain the sampling results of each water layer; wherein the sampling results include aquatic plant distribution data, aquatic animal distribution data and natural ecological environment data of the water layer; The database module is used to construct a target database based on the aquatic plant distribution data and the aquatic animal distribution data, calling the corresponding biological information and the natural ecological environment data of the water layer; The water area data dynamic change monitoring module is used to determine the dynamic change law of the target water area within a preset time according to the target database through the BP neural network; The plant community restoration calculation module is used to determine whether the dynamic change law meets the preset restoration goals. According to the judgment results and based on the principles of ecological restoration, the plant species and corresponding quantities required for each water layer are calculated to obtain the calculation results; the plant community management module is used to manage the plant community in the target water area using plant management equipment based on the calculation results.

Images