Fish and shrimp growth trend prediction method and system based on growth model

Abstract

The invention relates to a fish and shrimp growth trend prediction method and device based on a growth model. The method comprises the following steps: determining target fish and shrimp types, culture conditions and day age range data; obtaining weight data of the target fish and shrimp species in each day from the day age range data under the determined culture condition; preprocessing the weight data to obtain preprocessed weight data; determining a nonlinear model with good fitting efficiency in fishes and shrimps; based on the preprocessed weight data, performing modeling through a nonlinear model to obtain a first model, a second model and a third model; obtaining key information based on the first model, the second model and the third model; determining an optimal model in the first model, the second model and the third model based on the key information; and performing growth trend prediction on the target fish and shrimp species under the current culture condition by using the optimal model. The method can assist in determining the growth and development state of the animal, and brings great convenience to related workers.

Description

Technical Field

[0001] The present invention relates to the technical field related to smart farming, and in particular to a method and device for predicting the growth trends of fish and shrimp based on a growth model. Background Art

[0002] In the traditional model, when predicting the growth trend of fish and shrimp, the relevant staff often evaluate their own experience level and the obtained fish and shrimp parameters to make an assessment. There are problems such as evaluation reliance on experience, vague standards, and the reliability of judgment cannot be guaranteed. At present, there are many growth models for fish and shrimp that can predict the growth trend of fish and shrimp, but the applicability of growth models to different fish and shrimp species under different breeding conditions is different. As a result, it is difficult to quickly find growth models suitable for different fish and shrimp species, and then establish fish and shrimp growth trend prediction models in a targeted manner. Summary of the invention

[0003] The purpose of the present invention is to solve at least one of the deficiencies of the prior art and to provide a method and device for predicting the growth trend of fish and shrimp based on a growth model.

[0004] In order to achieve the above object, the present invention adopts the following technical solutions:

[0005] Specifically, a method for predicting the growth trend of fish and shrimp based on a growth model is proposed, including the following:

[0006] Determine the target fish and shrimp species, culture conditions and age range data;

[0007] Obtain the body weight data of the target fish and shrimp species for each day in the age range under the specified culture conditions;

[0008] Preprocessing the weight data to obtain preprocessed weight data;

[0009] Identify the nonlinear models Logistic, Gompertz and Von Bertalanffy models with good fitting performance in fish and shrimp;

[0010] Based on the preprocessed weight data, the nonlinear models Logistic, Gompertz and VonBertalanffy were used to respectively build the first model, the second model and the third model;

[0011] The key information is obtained based on the first model, the second model and the third model respectively, and the key information includes the limit weight, the age at the inflection point, the weight at the inflection point, the maximum daily weight gain, the goodness of fit R 2 Value and RMSE value;

[0012] Determine the optimal model among the first model, the second model and the third model based on the key information;

[0013] The optimal model is used to predict the growth trends of target fish and shrimp species under current farming conditions.

[0014] Further, specifically, the preprocessing operation includes:

[0015] The preprocessed weight data was obtained by removing outliers through data cleaning and then homogenizing.

[0016] Furthermore, SPSS20 software was used to iteratively estimate the parameters of the nonlinear models Logistic, Gompertz and VonBertalanffy models, and the goodness of fit R 2 The value and RMSE value were used as evaluation indicators, and the Levenberg-Marquardt algorithm was used to approximate the parameters. The square sum convergence threshold was set to 1E-8, and the parameter convergence threshold was set to 1E-8, and then the first model, the second model and the third model were obtained respectively.

[0017] Further, specifically, determining the optimal model among the first model, the second model and the third model based on the key information includes:

[0018] Determine the weight coefficient of each element in the key information through the hierarchical analysis method;

[0019] Convert each element in the key information into a value of the same type by normalization;

[0020] Calculate the sum of the values ​​converted from each element in the normalized key information in the first model, the second model, and the third model and multiply them by their corresponding weight coefficients to obtain a first value sum corresponding to the first model, a second value sum corresponding to the second model, and a third value sum corresponding to the third model;

[0021] Find the minimum value among the sum of the first value, the sum of the second value and the sum of the third value;

[0022] The model corresponding to the minimum value is taken as the optimal model.

[0023] The present invention also proposes a method for predicting the growth of largemouth bass, which applies the above-mentioned method for predicting the growth trend of fish and shrimp based on the growth model, comprising:

[0024] The target fish and shrimp species were determined to be largemouth bass, the culture conditions were large pond culture mode, and the age range data was 0 to 247 days old;

[0025] Finally, the optimal growth model was determined to be the Von Bertalanffy growth model, with the inflection point at day 145.89. The third model was found to be the optimal model, that is, W = 363.850 × (1-3.981 × e -0.017t ) 3 , where W represents body weight and t represents age in days;

[0026] In the large pond aquaculture mode, the optimal model is W = 363.850 × (1-3.981 × e -0.017t ) 3 Growth trend prediction for largemouth bass.

[0027] The present invention also proposes a method for predicting the growth of largemouth bass, which applies the above-mentioned method for predicting the growth trend of fish and shrimp based on the growth model, comprising:

[0028] The target fish and shrimp species were determined to be largemouth bass, the culture conditions were recirculating aquaculture mode, and the age range data was 0 to 273 days old;

[0029] The optimal growth model is determined to be the Von Bertalanffy growth model, with the inflection point at day 269.05. The third model is the optimal model, that is, W = 1081.795 × (1-1.2e -0.004761t ) 3 , where W represents body weight and t represents age in days;

[0030] In the recirculating aquaculture mode, the optimal model is W = 1081.795 × (1-1.2e -0.004761t ) 3 Growth trend prediction for largemouth bass.

[0031] The present invention also proposes a method for predicting the growth of Macrobrachium rosenbergii, which applies the above-mentioned method for predicting the growth trend of fish and shrimp based on the growth model, comprising:

[0032] The target fish and shrimp species were determined to be Macrobrachium rosenbergii, the culture conditions were the rice field culture mode, and the age range data was 0 to 192 days old;

[0033] The optimal growth model is determined to be the Logistic growth model, the inflection point age is 128.37 days, and the first model is the optimal model, that is, Where W represents body weight, and t represents age in days;

[0034] In the rice field farming mode, the optimal model is Predict the growth trend of Macrobrachium rosenbergii.

[0035] The present invention also proposes a fish and shrimp growth trend prediction device based on a growth model, comprising the following:

[0036] Condition setting module, used to determine the target fish and shrimp species, breeding conditions and age range data;

[0037] A data acquisition module is used to obtain the weight data of the target fish and shrimp species every day in the age range data under the determined breeding conditions;

[0038] A preprocessing module, used for preprocessing the weight data to obtain preprocessed weight data;

[0039] Model selection module, used to determine the nonlinear models with good fitting efficiency in fish and shrimp, such as Logistic, Gompertz and Von Bertalanffy models;

[0040] A model building module, for respectively building a first model, a second model and a third model based on the preprocessed weight data through nonlinear models Logistic, Gompertz and Von Bertalanffy models;

[0041] The key information acquisition module is used to obtain key information based on the first model, the second model and the third model respectively, wherein the key information includes the limit weight, the age at the inflection point, the weight at the inflection point, the maximum daily weight gain, the goodness of fit R 2 Value and RMSE value;

[0042] An optimal model calculation module, used to determine the optimal model among the first model, the second model and the third model based on the key information;

[0043] The prediction module is used to predict the growth trend of target fish and shrimp species under current breeding conditions using the optimal model.

[0044] The beneficial effects of the present invention are:

[0045] The present invention proposes a method and device for predicting the growth trend of fish and shrimp based on a growth model. After determining the target fish and shrimp species, breeding conditions and age range data, the present invention can find the most suitable nonlinear model for the current scenario from the nonlinear model with good fitting efficiency among fish and shrimp based on key information, and then use the modeling model to accurately predict the growth trend of the target fish and shrimp to assist in determining the growth and development status of the animals, bringing great convenience to relevant personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The above and other features of the present disclosure will become more apparent by describing in detail the embodiments shown in the accompanying drawings. The same reference numerals in the accompanying drawings of the present disclosure represent the same or similar elements. Obviously, the accompanying drawings described below are only some embodiments of the present disclosure. For those skilled in the art, other accompanying drawings can be obtained based on these accompanying drawings without creative work. In the accompanying drawings:

[0047] Figure 1 Shown is a flow chart of a method for predicting the growth trend of fish and shrimp based on a growth model of the present invention;

[0048] Figure 2-1 Shown is the iterative history of the Gompertz model fitted to the growth data of largemouth bass in large ponds;

[0049] Figure 2-2 The figure shows the growth model curve of the Gompertz model fitted to the growth data of largemouth bass cultured in large ponds;

[0050] Figure 2-3 Shown is the iterative history of the logistic model fitted to the growth data of largemouth bass in large ponds;

[0051] Figure 2-4 Shown is the growth model curve of the largemouth bass pond culture growth data fitted with the Logistic model;

[0052] Figure 2-5 Shown is the iterative history of the Von Bertalanffy model fitted to the growth data of largemouth bass in large ponds;

[0053] Figure 2-6 Shown is the growth model curve of the largemouth bass pond culture growth data fitted with the Von Bertalanffy model;

[0054] Figure 2-7 Shown is the iterative history of the Gompertz model fitted to the growth data of largemouth bass recirculating aquaculture;

[0055] Figure 2-8 Shown is the growth model curve of the Gompertz model fitted to the growth data of largemouth bass recirculating aquaculture;

[0056] Figure 2-9 Shown is the iterative history of the logistic model fitted to the growth data of largemouth bass recirculating aquaculture;

[0057] Figure 2-10 Shown is the growth model curve of the largemouth bass recirculating aquaculture growth data fitted with the Logistic model;

[0058] Figure 2-11 Shown is the iterative history of the Von Bertalanffy model fitted to the growth data of largemouth bass recirculating aquaculture;

[0059] Figure 2-12 Shown is the growth model curve of the largemouth bass recirculating aquaculture growth data fitted with the Von Bertalanffy model;

[0060] Figure 2-13 Shown is the iterative history of the Gompertz model fitted to the growth data of Macrobrachium rosenbergii in rice fields;

[0061] Figure 2-14 The figure shows the growth model curve of the Gompertz model fitted to the growth data of Macrobrachium rosenbergii cultured in rice fields;

[0062] Figure 2-15 Shown is the iterative history record in the logistic model fitted with the growth data of Macrobrachium rosenbergii in rice fields;

[0063] Figure 2-16 The figure shows the growth model curve of the logistic model fitted to the growth data of Macrobrachium rosenbergii cultured in rice fields;

[0064] Figure 2-17 The figure shows the iterative history of the Von Bertalanffy model fitted to the growth data of Macrobrachium rosenbergii in rice fields;

[0065] Figure 2-18 Shown is the growth model curve of the Von Bertalanffy model fitted to the growth data of Macrobrachium rosenbergii cultured in rice fields. DETAILED DESCRIPTION

[0066] The following will be combined with the embodiments and drawings to clearly and completely describe the concept, specific structure and technical effects of the present invention, so as to fully understand the purpose, scheme and effect of the present invention. It should be noted that the embodiments in this application and the features in the embodiments can be combined with each other without conflict. The same reference numerals used throughout the drawings indicate the same or similar parts.

[0067] Example 1, reference Figure 1 The present invention proposes a method for predicting the growth trend of fish and shrimp based on a growth model, comprising the following steps:

[0068] Step 110, determining target fish and shrimp species, breeding conditions, and age range data;

[0069] Step 120, obtaining the weight data of the target fish and shrimp species for each day in the age range data under the determined breeding conditions;

[0070] Step 130: preprocessing the weight data to obtain preprocessed weight data;

[0071] Step 140, determining the nonlinear models Logistic, Gompertz and Von Bertalanffy models that have good fitting performance in fish and shrimp;

[0072] Step 150: Based on the preprocessed weight data, nonlinear models Logistic, Gompertz and Von Bertalanffy are used to respectively model the first model, the second model and the third model;

[0073] Step 160: Obtain key information based on the first model, the second model, and the third model, respectively. The key information includes the limit weight, the age at the inflection point, the weight at the inflection point, the maximum daily weight gain, and the goodness of fit R 2 Value and RMSE value;

[0074] Step 170: determining the optimal model among the first model, the second model and the third model based on the key information;

[0075] Step 180: Use the optimal model to predict the growth trend of the target fish and shrimp species under the current breeding conditions.

[0076] In this embodiment 1, after determining the target fish and shrimp species, breeding conditions and age range data, the modeling model of the nonlinear model that is most suitable for the current scenario can be found from the nonlinear models with good fitting efficiency among fish and shrimp based on the key information, and then the growth trend of the target fish and shrimp can be accurately predicted through the modeling model to assist in determining the growth and development status of the animals, which brings great convenience to relevant personnel.

[0077] As a preferred embodiment of the present invention, specifically, the pretreatment operation includes:

[0078] The preprocessed weight data was obtained by removing outliers through data cleaning and then homogenizing.

[0079] As a preferred embodiment of the present invention, SPSS20 software is used to iteratively estimate the parameters of the nonlinear models Logistic, Gompertz and Von Bertalanffy models to obtain the goodness of fit R 2 The value and RMSE value were used as evaluation indicators, and the Levenberg-Marquardt algorithm was used to approximate the parameters. The square sum convergence threshold was set to 1E-8, and the parameter convergence threshold was set to 1E-8, and then the first model, the second model and the third model were obtained respectively.

[0080] As a preferred embodiment of the present invention, specifically, determining the optimal model among the first model, the second model and the third model based on the key information includes:

[0081] Determine the weight coefficient of each element in the key information through the hierarchical analysis method;

[0082] Convert each element in the key information into a value of the same type by normalization;

[0083] Calculate the sum of the values ​​converted from each element in the normalized key information in the first model, the second model, and the third model and multiply them by their corresponding weight coefficients to obtain a first value sum corresponding to the first model, a second value sum corresponding to the second model, and a third value sum corresponding to the third model;

[0084] Find the minimum value among the sum of the first value, the sum of the second value and the sum of the third value;

[0085] The model corresponding to the minimum value is taken as the optimal model.

[0086] In specific applications, in order to achieve accurate assessment of representative aquatic animals, it is necessary to combine body length detection with the reference growth model. Therefore, this chapter will study and construct the best animal growth model for representative aquatic animals under specific breeding conditions to assist in determining the growth and development status of animals.

[0087] 3.1 Experimental subjects and data collection

[0088] 3.1.1 Experimental subjects

[0089] Growth data were collected for representative aquatic animals, including largemouth bass cultured in large ponds (0 to 247 days old), recirculating aquaculture (0 to 273 days old), and Macrobrachium rosenbergii cultured in rice fields (0 to 192 days old). Logistic, Gompertz and Von Bertalanffy models were fitted respectively, and a total of 9 growth models were obtained.

[0090] 3.1.2 Data Collection

[0091] The data of largemouth bass and Macrobrachium rosenbergii were uniformly collected using a weighing instrument to measure the weight (g), and after data cleaning (removing outliers) and homogenization, SPSS20 software was used to build the model.

[0092] 3.2 Experimental model preliminary selection and explanation

[0093] At present, there is still controversy among scholars at home and abroad about the use of different growth and development models to fit the growth and development curves of different animals. Based on the current popular research on plant and animal growth models, and considering that the farmed animals being fitted are largemouth bass and Macrobrachium rosenbergii, this paper will study and preliminarily screen out the existing nonlinear models Logistic, Gompertz and Von Bertalanffy models with good fitting efficiency in fish and shrimp, to ensure that they can more accurately fit the representative aquatic animals largemouth bass and Macrobrachium rosenbergii, and then fit the growth data to obtain the best fitting curve.

[0094] 3.2.1 Explanation of the Gompertz model

[0095] 3.2.1.1 Model parameters

[0096] Table 3-1Gompertz Growth Model Parameters

[0097]

[0098] 3.2.1.2 Fitting parameters

[0099] The present invention uses SPSS20 software to iteratively estimate the parameters of the Gompertz model to calculate the goodness of fit R 2 As well as RMSE as the main evaluation indicators, the Levenberg-Marquardt algorithm is used to approximate the parameters, the sum of squares convergence threshold is 1E-8, and the parameter convergence threshold is 1E-8.

[0100] 3.2.2 Logistic model description

[0101] 3.2.2.1 Model parameters

[0102] Table 3-2Logistic Growth Model Parameters

[0103]

[0104] 3.2.2.2 Fitting parameters

[0105] The present invention uses SPSS20 software to iteratively estimate the parameters of the Logistic model to calculate the goodness of fit R 2 As well as RMSE as the main evaluation indicators, the Levenberg-Marquardt algorithm is used to approximate the parameters, the sum of squares convergence threshold is 1E-8, and the parameter convergence threshold is 1E-8.

[0106] 3.2.3 Explanation of Von Bertalanffy model

[0107] 3.2.3.1 Model parameters

[0108] Table 3-3 Von Bertalanffy Growth Model Parameters

[0109]

[0110] 3.2.3.2 Fitting parameters

[0111] The present invention uses SPSS20 software to iteratively estimate the parameters of the Von Bertalanffy model to calculate the goodness of fit R 2 As well as RMSE as the main evaluation indicators, the Levenberg-Marquardt algorithm is used to approximate the parameters, the sum of squares convergence threshold is 1E-8, and the parameter convergence threshold is 1E-8.

[0112] Embodiment 2, the present invention also proposes a method for predicting the growth of largemouth bass, which applies the above-mentioned method for predicting the growth trend of fish and shrimp based on the growth model, comprising:

[0113] The target fish and shrimp species were determined to be largemouth bass, the culture conditions were large pond culture mode, and the age range data was 0 to 247 days old;

[0114] Finally, the optimal growth model was determined to be the Von Bertalanffy growth model, with the inflection point at day 145.89. The third model was found to be the optimal model, that is, W = 363.850 × (1-3.981 × e -0.017t ) 3 , where W represents body weight and t represents age in days;

[0115] In the large pond aquaculture mode, the optimal model is W = 363.850 × (1-3.981 × e -0.017t ) 3 Growth trend prediction for largemouth bass.

[0116] In this embodiment 2,

[0117] 3.3.1 Fitting results of growth data of largemouth bass cultured in large ponds

[0118] 3.3.1.1 Results of fitting the Gompertz model to the growth data of largemouth bass cultured in large ponds

[0119] like Figure 2-1The optimal solution parameters were found after 19 main iterations of the Levenberg-Marquardt algorithm. The Gompertz model expression of largemouth bass cultured in Datang is:

[0120]

[0121] The model shows that its ultimate weight is 329.94 g, its age at inflection point is 149.88 days, its weight at inflection point reaches 121.78 g, its maximum daily weight gain is 3.03 g, and its R 2 is 0.999, RMSE is 3.422, and the model curve is as follows Figure 2-2 shown.

[0122] 3.3.1.2 Results of fitting logistic model to the growth data of largemouth bass cultured in large ponds

[0123] like Figure 2-3 The optimal solution parameters were found after 27 main iterations of the Levenberg-Marquardt algorithm. The Logistic model expression of largemouth bass cultured in Datang is:

[0124]

[0125] The model shows that its ultimate weight is 302.80 g, the inflection point is 157.47 days old, the weight at the inflection point is 151.40 g, the maximum daily weight gain is 4.01 g, and R 2 is 0.999, RMSE is 3.094, and the model curve is as follows Figure 2-4 shown.

[0126] 3.3.1.3 Results of fitting the Von Bertalanffy model to the growth data of largemouth bass cultured in large ponds

[0127] like Figure 2-5 The optimal solution parameters were found after 27 main iterations of the Levenberg-Marquardt algorithm. The Von Bertalanffy model expression of largemouth bass cultured in Datang is:

[0128] W = 363.850 × (1-3.981 × e -0.017t ) 3

[0129] The model shows that its ultimate weight is 363.80 g, its age at inflection point is 145.89 days, its weight at inflection point reaches 107.81 g, its maximum daily weight gain is 2.75 g, and its R 2 is 0.999, RMSE is 2.893, and the model curve is as follows Figure 2-6 shown.

[0130] Embodiment 3, the present invention also proposes a method for predicting the growth of largemouth bass, which applies the above-mentioned method for predicting the growth trend of fish and shrimp based on the growth model, comprising:

[0131] The target fish and shrimp species were determined to be largemouth bass, the culture conditions were recirculating aquaculture mode, and the age range data was 0 to 273 days old;

[0132] The optimal growth model is determined to be the Von Bertalanffy growth model, with the inflection point at day 269.05. The third model is the optimal model, that is, W = 1081.795 × (1-1.2e -0.004761t ) 3 , where W represents body weight and t represents age in days;

[0133] In the recirculating aquaculture mode, the optimal model is W = 1081.795 × (1-1.2e -0.004761t ) 3 Growth trend prediction for largemouth bass.

[0134] In this embodiment 3,

[0135] 3.3.2 Fitting results of growth data of largemouth bass in recirculating aquaculture

[0136] 3.3.2.1 Results of fitting the Gompertz model to the growth data of largemouth bass in recirculating aquaculture

[0137] like Figure 2-7 The optimal solution parameters were found after 33 main iterations of the Levenberg-Marquardt algorithm. The Gompertz model expression of largemouth bass in recirculating aquaculture is:

[0138]

[0139] The model shows that its ultimate weight is 567.10 g, its age at inflection point is 219.26 days, its weight at inflection point reaches 208.63 g, its maximum daily weight gain is 2.12 g, and its R 2 is 0.990, RMSE is 7.377, and the model curve is as follows Figure 2-8 shown.

[0140] 3.3.2.2 Results of fitting logistic model to the growth data of largemouth bass in recirculating aquaculture

[0141] like Figure 2-9 The optimal solution parameters were found after 40 main iterations of the Levenberg-Marquardt algorithm. The Logistic model expression of largemouth bass cultured in recirculating aquaculture is:

[0142]

[0143] The model shows that its ultimate weight is 334.75 g, its age at inflection point is 199.472 days, its weight at inflection point reaches 167.355 g, its maximum daily weight gain is 2.17 g, and its R 2 is 0.985, RMSE is 9.146, and the model curve is as follows Figure 2-10 shown.

[0144] 3.3.2.3 Results of fitting the Von Bertalanffy model to the growth data of largemouth bass in recirculating aquaculture

[0145] like Figure 2-11 The optimal solution parameters were found after 50 main iterations of the Levenberg-Marquardt algorithm. The Von Bertalanffy model expression of largemouth bass in recirculating aquaculture is:

[0146] W = 1081.795 × (1-1.2e -0.004761t ) 3

[0147] The model shows that its ultimate weight is 1081.80 g, the inflection point is 269.05 days old, the weight at the inflection point is 320.53 g, the maximum daily weight gain is 2.29 g, and R 2 is 0.992, RMSE is 6.516, and the model curve is as follows Figure 2-12 shown.

[0148] Embodiment 4, the present invention also proposes a method for predicting the growth of Macrobrachium rosenbergii, which applies the above-mentioned method for predicting the growth trend of fish and shrimp based on the growth model, comprising:

[0149] The target fish and shrimp species were determined to be Macrobrachium rosenbergii, the culture conditions were the rice field culture mode, and the age range data was 0 to 192 days old;

[0150] The optimal growth model is determined to be the Logistic growth model, the inflection point age is 128.37 days, and the first model is the optimal model, that is, Where W represents body weight, and t represents age in days;

[0151] In the rice field farming mode, the optimal model is Predict the growth trend of Macrobrachium rosenbergii.

[0152] In this embodiment 4,

[0153] 3.3.3 Fitting results of growth data of Macrobrachium rosenbergii cultured in rice fields

[0154] 3.3.3.1 Results of fitting the Gompertz model to the growth data of Macrobrachium rosenbergii in rice fields

[0155] like Figure 2-13 The optimal solution parameters were found after 36 main iterations of the Levenberg-Marquardt algorithm. The Gompertz model expression of Macrobrachium rosenbergii cultured in rice fields is:

[0156]

[0157] The model shows that its ultimate weight is 28.062 g, its inflection point is 111.15 days old, its weight reaches 10.32 g, and its maximum daily weight gain is 0.31 g. 2 is 0.986, RMSE is 0.761, and the model curve is as follows Figure 2-14 shown.

[0158] 3.3.3.2 Results of fitting logistic model to the growth data of Macrobrachium rosenbergii cultured in rice fields

[0159] like Figure 2-15 The optimal solution parameters were found after 45 main iterations of the Levenberg-Marquardt algorithm. The Logistic model expression of Macrobrachium rosenbergii cultured in rice fields is:

[0160]

[0161] The model shows that its ultimate body weight is 23.01 g, its inflection point is 128.37 days old, its body weight reaches 11.50 g, and its maximum daily weight gain is 0.33 g. 2 is 0.989, RMSE is 0.670, and the model curve is as follows Figure 2-16 shown.

[0162] 3.3.3.3 Results of fitting the Von Bertalanffy model to the growth data of Macrobrachium rosenbergii in rice fields

[0163] like Figure 2-17 The optimal solution parameters were found after 28 main iterations of the Levenberg-Marquardt algorithm. The Von Bertalanffy model expression of Macrobrachium rosenbergii cultured in rice fields is:

[0164] W = 33.418 × (1-3.806e -0.02t ) 3

[0165] The model shows that its ultimate weight is 33.42 g, its age at inflection point is 121.76 days, its weight at inflection point reaches 9.9 g, its maximum daily weight gain is 0.30 g, and R 2 is 0.989, RMSE is 0.686, and the model curve is as follows Figure 2-18 shown.

[0166] The experimental results of all growth model fitting in the above-mentioned Examples 2, 3 and 4 are shown in Tables 3-4 below:

[0167] Table 3-4 Growth model Parameters

[0168]

[0169] 3.4.2 Discussion on the screening of animal growth models

[0170] From the parameter analysis of various growth models of representative aquatic animals given in Table 3-4, the goodness of fit of the Gompertz model, Logistic model and Von Bertalanffy model fitting the growth data of largemouth bass under large pond farming conditions are all greater than 0.99, and the RMSE are 3.422, 3.094 and 2.893 respectively.

[0171] In order to further explore the optimal model for the growth data of largemouth bass cultured in large ponds, it is noted that the RMSE value of the VonBertalanffy model is the smallest. The size of this value directly reflects the prediction accuracy of the model, indicating that the VonBertalanffy model has higher prediction accuracy than the Gompertz model and the Logistic model, and can more accurately describe the growth performance of largemouth bass cultured in large ponds. The goodness of fit R2 of this model is also the highest. Therefore, the VonBertalanffy model is selected as the optimal growth model for largemouth bass cultured in large ponds, and the expression is:

[0172]

[0173] The inflection point age was 145.89 days, the inflection point weight was 107.81 grams, and the maximum daily weight gain was 2.75 grams. Under the conditions of recirculating aquaculture, the goodness of fit R2 of the three models fitting the growth data of largemouth bass were all greater than 0.999, and the RMSE were 7.377, 9.146 and 6.516 respectively. The Von Bertalanffy model also had the smallest RMSE value of 6.516, and the goodness of fit R2 of the model was also the highest. Therefore, the Von Bertalanffy model was selected as the optimal growth model for largemouth bass under recirculating water conditions, and the expression was:

[0174] W = 1081.795 × (1-1.2e -0.004761t ) 3

[0175] The inflection point age was 269.05 days, the inflection point weight was 320.53 g, and the maximum daily weight gain was 2.30 g. This is inconsistent with the best model obtained by He Xiaoyan et al.

[67] in their study of the early growth and development of largemouth bass. Their best model was the Logistic model, which may be due to the fact that it only fits the data of the early growth period and the different breeding environments; Tang Zhangsheng et al.

[61] and Li Jun et al.

[87] also used the three models included in the present invention to fit the growth data of GIFT tilapia (Oreochromis niloticus) and cage cultured lip fish bone (Hemibarbus labeo) respectively, and showed that the Von Bertalanffy model was the best model; El-Nabulsi et al.

[88] showed that the Von Bertalanffy equation can well describe the somatic cell growth after maturity, and it is also related to the growth of the whole individual. It is a model that describes mass balance and provides a reasonable description for the growth of most aquatic species. In the mathematical representation of the three model parameters, the symbol "A" represents the limit value (body weight) of the model expression. It is noted that the three models of the three different experimental groups all showed the fitting results of Von Bertalanffy>Gompertz>Logistic at the limit body weight A and the maximum daily weight gain. This is consistent with the research results of Li Jun et al.

[87] on the growth performance of cage-cultured bream bones using the same three growth models as the present invention. It also shows that Von Bertalanffy and Gompertz tend to give higher body weight estimates after fitting the growth data of representative aquatic animals.

[0176] Goodness of fit R of three models fitted to the growth data of Macrobrachium rosenbergii under the rice field culture mode 2 The RMSE values ​​of the Logistic model are all greater than 0.986, and the RMSE values ​​are 0.761, 0.670, and 0.686, respectively. The RMSE value of the Logistic model is the smallest, which is 6.516, and the goodness of fit of the model is R 2 The logistic model is also the highest, so the optimal growth model of Macrobrachium rosenbergii under the rice field culture mode is selected as the best model, and the expression is:

[0177]

[0178] The inflection point age was 128.37 days, the inflection point body weight was 11.50 g, and the maximum daily weight gain was 0.33 g. Yang Ming et al.

[89] used a multi-model method to estimate the growth parameters of Macrobrachium rosenbergii. The results showed that the application of Akaike Information Criterion (AIC) to the Logistic and Richards growth models showed that the optimal growth model for female shrimp was the Logistic model, which was consistent with the results of this experiment.

[0179] 3.4.3 Study on the characteristics of animal weight gain

[0180] Since it is difficult to distinguish the male and female of largemouth bass and Macrobrachium rosenbergii before sexual maturity, some scholars believe that the data differences in the growth process of male and female largemouth bass can be ignored. Considering that the growth model obtained in this chapter is to prepare for the application of the seedling counting and detection system in the actual environment, the present invention does not distinguish between male and female largemouth bass and Macrobrachium rosenbergii for curve fitting. The three growth models adopted in the present invention are theoretically used to describe the changes in objective indicators such as body weight of biological individuals during the entire growth process. The significant feature of the S-shaped curve is slow growth-rapid growth-slow growth period, and the derivative of the expression is a hump-shaped. From this perspective, the growth of plants and animals can be divided into four stages, which can generally be divided into a growth period, a growth acceleration period (the stage before the inflection point), a growth deceleration period after the derivative reaches the maximum value at the inflection point (the stage after the inflection point), and a growth stagnation period. The goodness of fit of the 9 models R 2 They are all very close to 1, indicating that the growth and development stages of largemouth bass and Macrobrachium rosenbergii basically conform to these four growth stages, and their growth rates change from slow to rapid and then to slow again.

[0181] It is worth noting that the weight inflection point represents the highest growth rate and time point of representative aquatic animals predicted by the model, after which the weight gain rate of farmed animals will gradually decrease. The weight of farmed animals will increase rapidly in the two stages before and after the inflection point, which greatly affects the economic benefits of farmers. The inflection point value given by the Von Bertalanffy model for pond farming of largemouth bass is 145.89 days. The optimal growth model in the study of the growth characteristics of largemouth bass under pond farming conditions by Xu Zhijie et al.

[90] showed that its inflection point age was 145 days and the body weight was 163.2 grams, which is very close to the inflection point age in the present invention. The difference in body weight may be due to differences in growth environment and feeding conditions.

[0182] In the present invention, the Von Bertalanffy model was used to culture largemouth bass in recirculating water, and the predicted inflection point age was 269.05 days, which was much longer than the 145.89 days of large pond culture. Zhou Xiuzhen et al.

[91] showed that the content of protein, ash, flavor amino acids, essential amino acids, and monounsaturated fatty acids increased after purification in the study of growth changes of largemouth bass in recirculating water purification culture. Yuan Julin et al.

[92] showed that appropriate system flow rate had a positive effect on the survival rate of largemouth bass in the study of fish growth in pond recirculating culture mode and conventional pond culture mode. Compared with the ordinary pond culture mode, the internal circulation culture mode can significantly improve the survival rate of fish, reduce the deposition of fish body fat, and promote protein synthesis. This may partially reveal the reason why the inflection point age and limit weight of largemouth bass cultured in recirculating water in the present invention are much higher than those in large pond culture.

[0183] In the present invention, the inflection point age predicted by the logistic model of Macrobrachium rosenbergii in rice field culture is 128.37 days. Yang Ming et al.

[93] only applied the Von Bertalanffy growth equation to fit the growth data of Macrobrachium rosenbergii in pond intensive culture, indicating that the growth inflection point of female shrimp is 111.16 days and the growth inflection point of male shrimp is 163.70 days, which indicates that the inflection point of the growth of Macrobrachium rosenbergii in rice field culture and pond intensive culture may not be greatly affected. The inflection point value can provide a reference for better large-scale farming and scientific feeding, reminding farmers to pay more attention to the transformation and implementation of the feeding strategy of breeding feed at this stage in order to obtain higher breeding economic benefits. On the other hand, it can also help farmers reduce water pollution caused by water quality deterioration due to feed deposition, damage to the health of farmed animals, and avoid economic losses caused by farmers.

[0184] In addition, through the present invention, accurate data show that under the condition of recirculating aquaculture, the corresponding limit body weights of the Gompertz model, the Logistic model and the Von Bertalanffy model of largemouth bass are 567.102 grams, 334.746 grams and 1081.795 grams respectively, which are higher than the values ​​of the limit body weights of the three models under the condition of large pond culture, 329.94 grams, 302.804 grams and 363.850 grams, especially the Von Bertalanff model is 197% higher, which means that under the condition of recirculating aquaculture, the limit growth body weight given by the growth model is much better than that of large pond culture, which may be due to the differences in water flow rate, population density, water index pH, water temperature, ammonia nitrogen, nitrite and feed feeding level under different environments, which greatly affects the development of the organism. Ido Seginer et al.

[94] studied the growth of gilthead sea bream (Sparus aurata) cultured in Mediterranean sea cages and recirculating aquaculture systems, describing them using the Von Bertalanffy model. Assuming that feed composition and environmental quality are no longer limiting factors, the analysis showed that the final individual size of recirculating aquaculture is much higher than that of sea cage culture, and the difference in feeding conditions is large. Appropriate temperature and stocking density are more conducive to feeding and healthy development of young fish. Alberto Beltrán et al.

[95] showed that although the initial investment of recirculating water systems is higher, they have more advantages, such as reducing unit production costs, improving water quality and improving feed conversion rate, which is more conducive to the healthy development of aquaculture animals.

[0185] So in summary,

[0186] (1) Under the large pond culture model, the optimal growth model of largemouth bass is the Von Bertalanffy growth model, with the inflection point at 145.89 days of age;

[0187] (2) Under the recirculating aquaculture mode, the optimal growth model of largemouth bass is the Von Bertalanffy growth model, and the inflection point is the 269.05th day.

[0188] (3) Under the rice field culture mode, the optimal growth model of Macrobrachium rosenbergii is the Logistic growth model, and the inflection point is the 128.37th day of age.

[0189] In summary, the above three growth models were selected as reference models for the seedling counting and growth evaluation detection system of largemouth bass and Macrobrachium rosenbergii under specific breeding conditions.

[0190] The present invention also proposes a fish and shrimp growth trend prediction device based on a growth model, comprising the following:

[0191] Condition setting module, used to determine the target fish and shrimp species, breeding conditions and age range data;

[0192] A data acquisition module is used to obtain the weight data of the target fish and shrimp species every day in the age range data under the determined breeding conditions;

[0193] A preprocessing module, used for preprocessing the weight data to obtain preprocessed weight data;

[0194] Model selection module, used to determine the nonlinear models with good fitting efficiency in fish and shrimp, such as Logistic, Gompertz and Von Bertalanffy models;

[0195] A model building module, for respectively building a first model, a second model and a third model based on the preprocessed weight data through nonlinear models Logistic, Gompertz and Von Bertalanffy models;

[0196] The key information acquisition module is used to obtain key information based on the first model, the second model and the third model respectively, wherein the key information includes the limit weight, the age at the inflection point, the weight at the inflection point, the maximum daily weight gain, the goodness of fit R 2 Value and RMSE value;

[0197] An optimal model calculation module, used to determine the optimal model among the first model, the second model and the third model based on the key information;

[0198] The prediction module is used to predict the growth trend of target fish and shrimp species under current breeding conditions using the optimal model.

[0199] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of software functional modules.

[0200] If the integrated module is implemented in the form of a software function module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or system that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electrical carrier signal, telecommunication signal and software distribution medium, etc.

[0201] Although the description of the present invention has been quite detailed and specifically describes several described embodiments, it is not intended to be limited to any of these details or embodiments or any particular embodiment, but should be regarded as providing a broad possible interpretation of these claims in view of the prior art by reference to the appended claims, thereby effectively covering the intended scope of the present invention. In addition, the above description of the present invention is based on the embodiments foreseeable by the inventor, and its purpose is to provide a useful description, and those non-substantial changes to the present invention that have not yet been foreseen may still represent equivalent changes to the present invention.

[0202] The above is only a preferred embodiment of the present invention. The present invention is not limited to the above implementation. As long as the technical effect of the present invention is achieved by the same means, it should belong to the protection scope of the present invention. Within the protection scope of the present invention, its technical scheme and / or implementation method can have various modifications and changes.

Claims

1. A method for predicting the growth trend of fish and shrimp based on a growth model, characterized in that: Includes the following: Determine the target fish and shrimp species, culture conditions and age range data; Obtain the body weight data of the target fish and shrimp species for each day in the age range under the specified culture conditions; Preprocessing the weight data to obtain preprocessed weight data; The nonlinear models Logistic, Gompertz and VonBertalanffy models with good fitting performance in fish and shrimp were determined; Based on the preprocessed weight data, the nonlinear models Logistic, Gompertz and Von Bertalanffy were used to respectively build the first model, the second model and the third model; The key information is obtained based on the first model, the second model and the third model respectively, and the key information includes the limit weight, the age at the inflection point, the weight at the inflection point, the maximum daily weight gain, the goodness of fit R 2 Value and RMSE value; Determine the optimal model among the first model, the second model and the third model based on the key information; The optimal model is used to predict the growth trends of target fish and shrimp species under current farming conditions.

2. The method for predicting the growth trend of fish and shrimp based on a growth model according to claim 1, characterized in that: Specifically, the preprocessing operation includes: The preprocessed weight data was obtained by removing outliers through data cleaning and then homogenizing.

3. The method for predicting the growth trend of fish and shrimp based on a growth model according to claim 1, characterized in that: Specifically, SPSS20 software was used to iteratively estimate the parameters of the nonlinear models Logistic, Gompertz and Von Bertalanffy models, and the goodness of fit R 2 The value and RMSE value were used as evaluation indicators, and the Levenberg-Marquardt algorithm was used to approximate the parameters. The square sum convergence threshold was set to 1E-8, and the parameter convergence threshold was set to 1E-8, and then the first model, the second model and the third model were obtained respectively.

4. The method for predicting the growth trend of fish and shrimp based on a growth model according to claim 1, characterized in that: Specifically, determining the optimal model among the first model, the second model, and the third model based on the key information includes: Determine the weight coefficient of each element in the key information through the hierarchical analysis method; Convert each element in the key information into a value of the same type by normalization; Calculate the sum of the values ​​converted from each element in the normalized key information in the first model, the second model, and the third model and multiply them by their corresponding weight coefficients to obtain a first value sum corresponding to the first model, a second value sum corresponding to the second model, and a third value sum corresponding to the third model; Find the minimum value among the sum of the first value, the sum of the second value and the sum of the third value; The model corresponding to the minimum value is taken as the optimal model.

5. A method for predicting the growth of largemouth bass, characterized in that: The method for predicting the growth trend of fish and shrimp based on the growth model as described in any one of claims 1 to 4 above is applied, comprising: The target fish and shrimp species were determined to be largemouth bass, the culture conditions were large pond culture mode, and the age range data was 0 to 247 days old; Finally, the optimal growth model was determined to be the Von Bertalanffy growth model, with the inflection point at day 145.

89. The third model was found to be the optimal model, that is, W = 363.850 × (1-3.981 × e -0.017t ) 3 , where W represents body weight and t represents age in days; In the large pond aquaculture mode, the optimal model is W = 363.850 × (1-3.981 × e -0.017t ) 3 Growth trend prediction for largemouth bass.

6. A method for predicting the growth of largemouth bass, characterized in that: The method for predicting the growth trend of fish and shrimp based on the growth model as described in any one of claims 1 to 4 above is applied, comprising: The target fish and shrimp species were determined to be largemouth bass, the culture conditions were recirculating aquaculture mode, and the age range data was 0 to 273 days old; The optimal growth model is determined to be the Von Bertalanffy growth model, with the inflection point at day 269.

05. The third model is the optimal model, that is, W = 1081.795 × (1-1.2e -0.004761t ) 3 , where W represents body weight and t represents age in days; In the recirculating aquaculture mode, the optimal model is W = 1081.795 × (1-1.2e -0.004761t ) 3 Growth trend prediction for largemouth bass.

7. A method for predicting the growth of Macrobrachium rosenbergii, characterized in that: The method for predicting the growth trend of fish and shrimp based on the growth model as described in any one of claims 1 to 4 above is applied, comprising: The target fish and shrimp species were determined to be Macrobrachium rosenbergii, the culture conditions were the rice field culture mode, and the age range data was 0 to 192 days old; The optimal growth model is determined to be the Logistic growth model, the inflection point age is 128.37 days, and the first model is the optimal model, that is, Where W represents body weight, and t represents age in days; In the rice field farming mode, the optimal model is Predict the growth trend of Macrobrachium rosenbergii.

8. A device for predicting the growth trend of fish and shrimp based on a growth model, characterized in that: Includes the following: Condition setting module, used to determine the target fish and shrimp species, breeding conditions and age range data; A data acquisition module is used to obtain the weight data of the target fish and shrimp species every day in the age range data under the determined breeding conditions; A preprocessing module, used for preprocessing the weight data to obtain preprocessed weight data; Model selection module, used to determine the nonlinear models with good fitting efficiency in fish and shrimp, such as Logistic, Gompertz and Von Bertalanffy models; A model building module, for respectively building a first model, a second model and a third model based on the preprocessed weight data through nonlinear models Logistic, Gompertz and Von Bertalanffy models; The key information acquisition module is used to obtain key information based on the first model, the second model and the third model respectively, wherein the key information includes the limit weight, the age at the inflection point, the weight at the inflection point, the maximum daily weight gain, the goodness of fit R 2 Value and RMSE value; An optimal model calculation module, used to determine the optimal model among the first model, the second model and the third model based on the key information; The prediction module is used to predict the growth trend of target fish and shrimp species under current breeding conditions using the optimal model.

Images