Intelligent digital monitoring and management system for toad breeding based on the Internet of Things

Abstract

The present invention relates to the field of monitoring technology, and more specifically to an intelligent digital monitoring and management system for toad farming based on the Internet of Things. The system comprises: using an infrared sensor array arranged in a ground area to collect the number of toads in multiple ground sub-areas within the ground area to obtain a toad population distribution; collecting the ventilation frequency of multiple water sub-areas within a water surface area to obtain a ventilation frequency distribution; predicting the progress of gill degeneration based on the toad population distribution and ventilation frequency distribution to obtain a gill degeneration progress distribution and multiple confidence levels; obtaining an ideal ground humidity distribution and an ideal dissolved oxygen distribution; collecting the actual humidity distribution of multiple ground sub-areas and the actual dissolved oxygen distribution of multiple water sub-areas, calculating the fitness of farming parameters for the ideal ground humidity distribution and the ideal dissolved oxygen distribution, obtaining the fitness of farming parameters, and conducting toad farming monitoring and management. The system achieves the technical effect of timely adjusting the farming environment to the optimal environment for toad farming based on the growth status of the toads.

Description

Technical Field

[0001] The present invention relates to the field of monitoring technology, and in particular to an intelligent digital monitoring and management system for toad breeding based on the Internet of Things. Background Art

[0002] Toad farming is a vital component of both medicinal and ecological economics. Accurate monitoring of the toad's growth environment and physiological status directly impacts the toad's health, the accumulation of medicinal ingredients (such as toad venom), and the profitability of the farming process. Traditional farming methods rely primarily on empirical assessment of environmental parameters (such as humidity and dissolved oxygen) and physiological indicators (such as the progression of gill degeneration). This approach suffers from a single environmental monitoring method and inaccurate assessments of the toad's growth status. Furthermore, there is a technical limitation that prevents timely adjustment of the farming environment to the optimal conditions for toad cultivation based on the toad's growth status. Summary of the Invention

[0003] The present invention aims to solve the technical problem in the prior art that the breeding environment cannot be adjusted to the optimal environment for toad breeding in time according to the growth status of toads, and provides an intelligent digital monitoring and management system for toad breeding based on the Internet of Things.

[0004] The technical solution of the present invention to solve the above technical problems is as follows:

[0005] The present invention provides an intelligent digital monitoring and management system for toad farming based on the Internet of Things, comprising:

[0006] A toad information collection unit is used to collect the number of toads in multiple ground zones in the ground area through an infrared sensor array deployed in the ground area based on the Internet of Things, thereby obtaining the toad population distribution; and to collect the ventilation frequency of multiple water zones in the water area through an infrared sensor array deployed in the water area, thereby obtaining the ventilation frequency distribution;

[0007] a gill degeneration degree prediction unit, configured to predict the gill degeneration progress according to the toad population distribution and ventilation frequency distribution, and obtain the gill degeneration progress distribution and a plurality of confidence levels;

[0008] An environmental parameter indexing unit is used to perform ideal ground humidity indexing and ideal dissolved oxygen indexing according to the gill degeneration progress distribution, so as to obtain an ideal ground humidity distribution and an ideal dissolved oxygen distribution;

[0009] The breeding parameter adaptation unit is used to collect the actual humidity distribution of the multiple ground partitions and the actual dissolved oxygen distribution of the multiple water partitions based on the Internet of Things, and calculate the breeding parameter adaptation of the ideal ground humidity distribution and the ideal dissolved oxygen distribution in combination with the multiple confidence levels to obtain the breeding parameter adaptation and perform toad breeding monitoring and management.

[0010] Optionally, the toad information collection unit is also used to: based on the Internet of Things, collect ground infrared sensor data of multiple ground partitions in the ground area through an infrared sensor array arranged in the ground area; identify and obtain multiple toad numbers based on the multiple ground infrared sensor data, and obtain the toad number distribution; based on the Internet of Things, collect water area infrared sensor data of multiple water area partitions in the water surface area within a preset time period through an infrared sensor array arranged in the water surface area; identify and obtain multiple ventilation numbers based on the multiple water area infrared sensor data, and calculate and obtain multiple ventilation frequencies as the ventilation frequency distribution.

[0011] Optionally, the gill degeneration degree prediction unit is also used to: obtain the breeding quantity of the current batch of toads, and calculate the non-degenerated quantity in combination with the toad quantity distribution; input the multiple toad quantities within the toad quantity distribution into the ground gill degeneration prediction network, and predict and output multiple ground gill degeneration progress; combine the non-degenerated quantity with the multiple ventilation frequencies within the ventilation frequency distribution, and input the water area gill degeneration prediction network, and predict and output multiple water area gill degeneration progress; combine the multiple ground gill degeneration progress and the multiple water area gill degeneration progress to obtain a gill degeneration progress distribution; and process and obtain multiple confidence levels according to the gill degeneration progress distribution.

[0012] Among them, the training steps of the terrestrial gill degeneration prediction network and the aquatic gill degeneration prediction network include: based on the gill degeneration observation data of toad farming, collecting a set of sample toad numbers in the terrestrial partition, as well as collecting a set of sample non-degenerated numbers and a set of sample ventilation frequencies in the aquatic partition; based on the gill degeneration observation data of toad farming, collecting and labeling a set of sample gill degeneration progress; using machine learning to construct a terrestrial gill degeneration prediction network and an aquatic gill degeneration prediction network; using the sample gill degeneration progress set as output supervision data and the sample toad number set as input training data, the terrestrial gill degeneration prediction network is supervised and trained until convergence; using the sample gill degeneration progress set as output supervision data and the sample non-degenerated number set and the sample ventilation frequency set in the aquatic partition as input training data, the aquatic gill degeneration prediction network is supervised and trained until convergence.

[0013] Among them, according to the gill degeneration progress distribution, multiple confidence levels are obtained by processing, including: calculating the error margin of each gill degeneration progress and the mean of the gill degeneration progress distribution, and calculating multiple first confidence levels; calculating the ratio of each toad number to the mean of the toad number distribution, and obtaining multiple second ground partition confidence levels; calculating the ratio of each ventilation frequency to the mean of the ventilation frequency distribution, and obtaining multiple second water partition confidence levels; according to the first confidence level and the second ground partition confidence level of each ground partition, as well as the first confidence level and the second water partition confidence level of each water partition, multiple confidence levels are calculated.

[0014] Optionally, the environmental parameter indexing unit is also used to: perform ideal ground humidity indexing and ideal dissolved oxygen indexing according to the gill degeneration progress distribution, and obtain an ideal ground humidity distribution and an ideal dissolved oxygen distribution, including: inputting the gill degeneration progress corresponding to the first ground partition within the gill degeneration progress distribution into a first ground humidity index table, and indexing to obtain a first ideal ground humidity, wherein the first ground humidity index table includes a mapping pair of sample ideal humidity and sample gill degeneration progress of the first ground partition; continuing to index to obtain multiple ideal ground humidity according to the gill degeneration progress corresponding to multiple ground partitions, and obtain an ideal ground humidity distribution; inputting the gill degeneration progress corresponding to the first water partition within the gill degeneration progress distribution into a first dissolved oxygen index table, and indexing to obtain a first ideal dissolved oxygen, wherein the first dissolved oxygen index table includes a mapping pair of sample ideal dissolved oxygen and sample gill degeneration progress of the first water partition; continuing to index to obtain multiple ideal dissolved oxygen according to the gill degeneration progress corresponding to multiple water partitions, and obtain an ideal dissolved oxygen distribution.

[0015] Optionally, the breeding parameter adaptation unit is also used to: based on the Internet of Things, collect the actual humidity distribution of the multiple ground partitions and the actual dissolved oxygen distribution of the multiple water partitions, combine the multiple confidence levels, calculate the breeding parameter fitness of the ideal ground humidity distribution and the ideal dissolved oxygen distribution, and obtain the breeding parameter fitness; based on the Internet of Things, collect the actual humidity of the multiple ground partitions to obtain the actual humidity distribution, collect the actual dissolved oxygen of the multiple water partitions to obtain the actual dissolved oxygen distribution; calculate the deviation amplitude of the actual humidity distribution from the ideal ground humidity distribution, and calculate the fitness to obtain the humidity fitness distribution; calculate the deviation amplitude of the actual dissolved oxygen distribution from the ideal dissolved oxygen distribution, and calculate the fitness to obtain the dissolved oxygen fitness distribution; according to the multiple confidence levels, perform weighted calculation on the fitness within the humidity fitness distribution and the dissolved oxygen fitness distribution to obtain the breeding parameter fitness, and perform toad breeding monitoring and management.

[0016] By implementing the present invention, it is possible to realize, based on the Internet of Things, using an infrared sensor array deployed in a ground area, collecting the number of toads in multiple ground partitions in the ground area and obtaining the toad population distribution; and using an infrared sensor array deployed in a water area, collecting the ventilation frequency of multiple water partitions in the water area and obtaining the ventilation frequency distribution, thereby accurately collecting growth-related parameters of toads in water and ground areas, and providing a scientific basis for predicting the progress of toad gill degeneration;

[0017] By implementing the present invention, it is possible to predict the progress of gill degeneration based on the toad number distribution and ventilation frequency distribution, obtain the gill degeneration progress distribution and multiple confidence levels, and accurately predict the progress of toad gill degeneration based on toad growth-related parameters, thereby improving the prediction efficiency and eliminating the randomness of manual prediction.

[0018] By implementing the present invention, it is possible to respectively perform ideal ground humidity indexing and ideal dissolved oxygen indexing according to the gill degeneration progress distribution, obtain ideal ground humidity distribution and ideal dissolved oxygen distribution, and determine the environmental parameters most suitable for the toad's growth at the current stage.

[0019] By implementing the present invention, it is possible to collect the actual humidity distribution of the multiple ground partitions and the actual dissolved oxygen distribution of the multiple water partitions based on the Internet of Things, combine the multiple confidence levels, calculate the breeding parameter adaptability of the ideal ground humidity distribution and the ideal dissolved oxygen distribution, obtain the breeding parameter adaptability, and conduct toad breeding monitoring and management, thereby achieving the matching and adjustment of the ideal ground humidity distribution and the ideal dissolved oxygen distribution that are most suitable for the current growth of the toad according to the current degree of gill degeneration of the toad, so as to facilitate the growth of the toad.

[0020] In summary, by implementing the present invention, the technical effect of timely adjusting the breeding environment to the optimal environment for toad breeding according to the growth conditions of toads can be achieved to facilitate the growth of toads. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic structural diagram of an intelligent digital monitoring and management system for toad farming based on the Internet of Things provided by the present invention;

[0022] Figure 2 This is a schematic diagram of the execution steps of a gill degeneration degree prediction unit in an intelligent digital monitoring and management system for toad farming based on the Internet of Things provided by the present invention.

[0023] In the accompanying drawings, the components represented by the reference numerals are as follows:

[0024] Toad information collection unit 11, gill degeneration degree prediction unit 12, environmental parameter indexing unit 13, and breeding parameter adaptation unit 14. DETAILED DESCRIPTION

[0025] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making any creative efforts shall fall within the scope of protection of the present invention.

[0026] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the specified features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0027] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or illustration". Any embodiment of the present invention described as "for example" is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any person skilled in the art to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes are not elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed herein.

[0028] Example 1, as Figure 1 As shown, the embodiment of the present invention provides an intelligent digital monitoring and management system for toad farming based on the Internet of Things, including:

[0029] The toad information collection unit 11 is used to collect the number of toads in multiple ground zones in the ground area through an infrared sensor array arranged in the ground area based on the Internet of Things, and obtain the toad population distribution; and to collect the ventilation frequency of multiple water zones in the water area through an infrared sensor array arranged in the water area, and obtain the ventilation frequency distribution;

[0030] a gill degeneration degree prediction unit 12 for predicting the gill degeneration progress according to the toad population distribution and the ventilation frequency distribution, and obtaining the gill degeneration progress distribution and a plurality of confidence levels;

[0031] An environmental parameter indexing unit 13 is used to perform ideal ground humidity indexing and ideal dissolved oxygen indexing according to the gill degeneration progress distribution, and obtain an ideal ground humidity distribution and an ideal dissolved oxygen distribution;

[0032] The breeding parameter adaptation unit 14 is used to collect the actual humidity distribution of the multiple ground partitions and the actual dissolved oxygen distribution of the multiple water partitions based on the Internet of Things, and calculate the breeding parameter adaptation of the ideal ground humidity distribution and the ideal dissolved oxygen distribution in combination with the multiple confidence levels to obtain the breeding parameter adaptation and perform toad breeding monitoring and management.

[0033] In the embodiment of the present application, the toad information collection unit 11 is based on the Internet of Things. By using an infrared sensor array arranged in the ground area, the number of toads in multiple ground partitions in the ground area is collected to obtain the toad population distribution. By using an infrared sensor array arranged in the water area, the ventilation frequency of multiple water area partitions in the water area is collected to obtain the ventilation frequency distribution, including:

[0034] Based on the Internet of Things, an infrared sensor array is deployed in the ground area to collect ground infrared sensing data of multiple ground partitions in the ground area;

[0035] Based on multiple ground infrared sensor data, multiple toad populations are identified and their distribution is obtained.

[0036] Based on the Internet of Things, an infrared sensor array is deployed on the water surface area to collect water area infrared sensing data of multiple water areas within a preset time period in the water surface area;

[0037] Based on infrared sensing data from multiple water areas, multiple ventilation quantities are identified and obtained, and multiple ventilation frequencies are calculated to obtain ventilation frequency distribution.

[0038] In the embodiment of the present application, the number of toads in the ground partition and the number of toads breathing in the water area are collected in order to predict the degree of gill degeneration of the toads based on the number of toads that have come ashore and the ventilation frequency of toads in the water, so as to determine the growth stage of the toads and adjust the environmental parameters such as humidity and dissolved oxygen for the toads according to their growth stage to facilitate their growth.

[0039] In an embodiment of the present application, the distribution of toad populations in a ground area is obtained by collecting ground infrared sensor data from multiple ground subareas within the ground area using an infrared sensor array disposed within the ground area. The infrared sensor array comprises multiple infrared sensors distributed throughout the ground area. The aforementioned ground area refers to the land area (such as mudflats, bushes, grasslands, etc.) within the toad breeding area excluding water. The ground subareas are areas divided by area for ease of management and calculation, such as 10 square meters per subarea, with the ground area being divided into several equally sized subareas. Optionally, collecting ground infrared sensor data from multiple subareas within the ground area can utilize a pyroelectric infrared sensor (PIR) or an infrared thermal imaging camera, such as the Hikvision DS-2TD series, to monitor the toad heat signatures and calculate the population distribution. The aforementioned distribution can be the number of toads within each subarea, such as 28 toads within the first subarea, 30 toads within the second subarea, and so on. The total number of toads within the ground area can be determined by summing the number of toads within all subareas.

[0040] Furthermore, the distribution parameters of toad breathing frequency in the water surface area can be collected by using an infrared sensor array installed in the water surface area to collect infrared sensor data from multiple water zones within the water surface area. Specifically, the infrared sensor array installed in the water surface area can first determine the number of toads in the water area, and then use infrared sensor images to capture the toads surfacing to breathe. Specifically, a waterproof infrared sensor (such as the FLIR TG165-X) can be used to capture the toads surfacing to breathe. The water zones can also be divided into areas that are easier to manage and calculate, such as 10 square meters per zone, and the water surface area can be divided into several equally sized zones. The toad breathing frequency in the water surface area is the number of toads breathing per unit time collected in a specific zone, such as 40 times per hour.

[0041] In the embodiment of the present application, the gill degeneration degree prediction unit 12 is used to predict the gill degeneration progress according to the toad number distribution and the ventilation frequency distribution, and obtain the gill degeneration progress distribution and multiple confidence levels, such as Figure 2 As shown, the gill degeneration degree prediction unit 12 includes the following execution steps:

[0042] S100: Obtain the number of toads bred in the current batch, and calculate the number of toads that have not degenerated based on the distribution of the number of toads;

[0043] S200: Inputting multiple toad numbers within the toad number distribution into a ground gill degeneration prediction network, and outputting predictions to obtain multiple ground gill degeneration progresses; inputting the non-degenerated numbers into a water area gill degeneration prediction network in combination with multiple ventilation frequencies within the ventilation frequency distribution, and outputting predictions to obtain multiple water area gill degeneration progresses;

[0044] S300: combining the plurality of terrestrial gill degeneration progresses and the plurality of water gill degeneration progresses to obtain a gill degeneration progress distribution;

[0045] S400: Processing to obtain multiple confidence levels according to the gill degeneration progress distribution.

[0046] In the step S100 of the present embodiment, the number of toads raised in the current batch of toads is obtained by combining the distribution of the toad population to calculate the number of toads whose gills have not completely degenerated. Toads with pre-degenerated gills primarily move about in the water. Therefore, the number of toads in the water area can be calculated based on the difference between the number of toads in the ground area and the total number of toads in the breeding ground. The number of toads in the water area can be calculated by subtracting the total number of toads in the ground area from the total number of toads in the breeding ground. For example, if the number of toads raised in the current batch of toads in the breeding ground is 1,000, and the total number of toads in the ground area obtained by the toad information collection unit 11 is 300, then the number of toads in the ground area = the total number of toads in the breeding ground (the number of toads raised in the current batch of toads) - the number of toads in the ground area = 1,000 - 300 = 700. The number of toads raised in the current batch of toads is the number of toad larvae released in the early stage.

[0047] The training steps of the gill degeneration prediction network and the water area gill degeneration prediction network described in step S200 of the embodiment of the present application include:

[0048] Based on the gill degeneration observation data of toad farming, the number of sample toads in the ground partition was collected, as well as the number of samples without degeneration and the ventilation frequency of samples in the water partition were collected;

[0049] Based on the gill degeneration observation data of toad farming, the gill degeneration progress set of samples was collected and annotated;

[0050] Using machine learning, we constructed a terrestrial gill degeneration prediction network and an aquatic gill degeneration prediction network;

[0051] The sample gill degeneration progress set is used as output supervision data, and the sample toad quantity set is used as input training data to perform supervised training on the terrestrial gill degeneration prediction network until convergence;

[0052] The sample gill degradation progress set is used as output supervision data, the sample non-degraded quantity set and the sample ventilation frequency set in the water area partition are used as input training data, and the water area gill degradation prediction network is supervised and trained until convergence.

[0053] In the embodiment of the present application, the ground gill degeneration prediction network and the water gill degeneration prediction network are trained using the sample toad number set in the ground partition, the sample non-degenerated number set and the sample ventilation frequency set in the water partition, respectively.

[0054] The terrestrial gill degeneration prediction network uses the sample gill degeneration progress set as output supervision data and the sample toad number set as input training data. After the training is completed, the toad number set is input to output the predicted gill degeneration progress.

[0055] The water area gill degeneration prediction network uses the sample gill degeneration progress set as the output supervision data, and the sample non-degraded number set and the sample ventilation frequency set in the water area partition as the input training data. After the training is completed, the non-degraded number and the sample ventilation frequency in the water area partition are input to predict the output gill degeneration progress.

[0056] The gill degeneration rate is the ratio of toads that have completed gill degeneration to the total number of toads in the current batch. For example, if 700 toads have completed gill degeneration and the total number of toads in the current batch is 1000, then the gill degeneration rate = 700 / 1000 = 0.7 = 70%. In this embodiment of the present application, to associate the degree of gill degeneration of toads with the aforementioned parameters such as the number of toads, the number of non-degenerated toads, and the ventilation rate, it is first necessary to collect gill degeneration rates under different conditions: the number of toads, the number of non-degenerated toads, and the ventilation rate.

[0057] The gill degeneration progress rate is the ratio of toads with complete gill degeneration to the total number of toads in the current batch. Determining whether a toad has completed gill degeneration requires manual testing. This testing method involves direct observation of the gill area on both sides of the toad's head. In juvenile toads (tadpoles), the external gills are typically red, feather-like structures located at the back of the head. After metamorphosis is complete (i.e., gill degeneration is complete), the external gills should completely disappear, leaving only traces of gill slits covered by the skin. After each test, the gill degeneration progress rate for the current batch of toads is recorded. Record the gill degeneration progress of multiple toad farms (for example, record it daily during the toad farming cycle, marking the time of each recording) to obtain a set of sample gill degeneration progress. Each time the sample gill degeneration progress is recorded, simultaneously obtain the number of toads in each partition at the time of gill degeneration testing, the number of toads that have not degenerated, and the ventilation frequency. Collect at least 200 sets of time-series samples (covering at least three farming batches) using this method to obtain a set of sample number of toads, a set of sample number of toads that have not degenerated, and a set of sample ventilation frequencies. This data set can be divided into a training set and a validation set in a ratio of 7:3 to train the terrestrial and aquatic gill degeneration prediction networks. Alternatively, randomly sample individuals from the same batch for testing (e.g., 5-10 per 100 individuals). Combined with the above method, this method can be used to rapidly assess overall developmental progress, reducing the workload of individual testing.

[0058] Alternatively, a spatiotemporal separation residual convolutional network (ST-ResNet) can be used to construct a gill degeneration prediction network. This network consists of an input layer, a spatiotemporal feature separation module, a residual stacking module, and an output layer. The input layer contains the number of toads detected in each partition (e.g., daily detection). The spatiotemporal feature separation module includes a temporal convolution layer and a spatial attention layer. The temporal convolution layer uses one-dimensional dilated convolution to extract short- and long-term temporal features, has 32 output channels, and uses ReLU as the activation function. The spatial attention layer uses multi-head self-attention with 4 heads. The residual stacking module consists of four groups of residual blocks, each consisting of a BatchNorm layer, a temporal convolution layer, a spatial convolution layer, and skip connections. The convolution kernels are 3×1 in temporal and 1×3 in spatial, with increasing numbers of channels (from 32, 64, to 128). The activation function is ELU. The output layer is a fully connected layer with a sigmoid activation function.

[0059] In the parameter settings of the terrestrial gill degeneration prediction network, the initial learning rate value is 0.005, and cyclic learning rate scheduling is used; the training batch size is 64; the optimizer uses NAdam; an inter-layer Dropout of 0.25 is added to prevent overfitting; the loss function uses quantile loss; the model evaluation metric is MAE, that is, the prediction error of the single partition data. When MAE is less than or equal to 2.5%, the model is judged to have converged. A terrestrial gill degeneration prediction network is obtained. By inputting the number of toads in a certain terrestrial partition, the gill degeneration progress corresponding to the number of toads in the corresponding terrestrial partition can be predicted. By inputting the number of toads within the toad population distribution into the terrestrial gill degeneration prediction network, multiple terrestrial gill degeneration progress predictions can be obtained.

[0060] Optionally, for the water gill degeneration prediction network, since the input features are the number of non-degraded gills and the ventilation frequency, the construction method is different from the terrestrial gill degeneration prediction network. Specifically, a nonlinear mapping model can be constructed using a dual-channel residual fully connected network (Dual-ResNet), including an input layer, a dual-channel feature extraction layer, a feature fusion layer, and an output layer.

[0061] Specifically, its input feature dimension is 2, channel 1 is the ventilation frequency (e.g., 40 times / h), and channel 2 is the number of non-degraded (e.g., 70%);

[0062] The ventilation frequency channel in the dual-channel feature extraction layer consists of a fully connected layer (64 neurons, using the ReLU activation function), followed by a residual block (including batch normalization, a fully connected layer (64 neurons), and the ReLU activation function), and finally a dropout layer with a dropout rate of 0.2. The non-degraded number channel consists of a fully connected layer (32 neurons, using the LeakyReLU activation function), followed by a residual block (including layer normalization, a fully connected layer (32 neurons), and the ELU activation function).

[0063] The feature fusion layer is used to concatenate the output of the ventilation frequency channel (64 dimensions) and the output of the non-degraded number channel (32 dimensions) in the feature dimension to generate a 96-dimensional joint feature vector.

[0064] The output layer is a fully connected layer. The first fully connected layer contains 32 neurons and uses the Swish activation function. The second fully connected layer contains 1 neuron and uses the Sigmoid activation function to limit the output value between 0 and 1, corresponding to 0% to 100% gill degeneration progress.

[0065] In training the gill degradation prediction network, the learning rate was set to an initial value of 0.002; the AdamW optimizer was used, with weight decay set to 0.005, β1=0.9, and β2=0.999; the training batch size was 16 to 32; the output layer activation function was Sigmoid; and the regularization for channel 1 was L2=0.001. Huber Loss (δ=0.5) was used as the main loss function during training. Early stopping was triggered when the validation set MAE did not decrease for 20 consecutive rounds. The model was considered converged when the MAE of the predicted results was ≤2.5%. The gill degradation prediction network was obtained. The number of non-degraded samples and the ventilation frequency of samples within a specific water area were input to predict the gill degradation progress. The non-degraded samples were combined with multiple ventilation frequencies within the ventilation frequency distribution and fed into the gill degradation prediction network to predict the gill degradation progress of multiple water areas.

[0066] Among them, in step S300 of the embodiment of the present application, the multiple ground gill degeneration progress and the multiple water gill degeneration progress are combined to obtain the gill degeneration progress distribution. Specifically, the aforementioned multiple ground gill degeneration progress and the multiple water gills are arranged one by one according to the corresponding ground partitions and water partitions to obtain the gill degeneration progress distribution.

[0067] The step S400 of the embodiment of the present application, in which the plurality of confidence levels are obtained according to the gill degeneration progress distribution, includes:

[0068] Calculating the error margin of each gill degeneration progress and the mean of the gill degeneration progress distribution, and calculating and obtaining a plurality of first confidence levels;

[0069] Calculate the ratio of each toad number to the mean of the toad number distribution to obtain multiple second ground partition confidences;

[0070] Calculating the ratio of each ventilation frequency to the mean of the ventilation frequency distribution to obtain a plurality of second water area partition confidences;

[0071] A plurality of confidence levels are calculated based on the first confidence level and the second confidence level of each ground partition, and the first confidence level and the second confidence level of each water partition.

[0072] In an embodiment of the present application, to calculate the first confidence level, it is first necessary to calculate the mean (e.g., 50%) of all gill degeneration progress values ​​in the gill degeneration progress distribution. The deviation between the value of a gill degeneration progress value in the gill degeneration progress distribution (e.g., 80%) and the aforementioned mean value can be (80% to 50%) / 50%=0.6. Optionally, the first confidence level can be the inverse of this deviation, i.e., first confidence level = 1 / deviation level = 1 / 0.6 = 1.667. The larger the first confidence level, the higher the credibility of the corresponding gill degeneration progress. According to the above method, the error range between each gill degeneration progress value and the mean of the gill degeneration progress distribution is calculated, and multiple first confidence levels can be calculated accordingly.

[0073] In this embodiment of the present application, to obtain the second ground partition confidence level, one first calculates the ratio of the number of toads in a particular ground partition (e.g., 27) to the mean number of toads in all ground partitions within the ground area (e.g., 30) (e.g., 18 / 30 = 0.6). The reciprocal of the absolute value of the difference between this ratio and 1 is then calculated as the second ground partition confidence level. For example, in the above example, the second ground partition confidence level = 1 / (|0.6-1|) = 2.5. By calculating the ratio of each toad number to the mean of the toad number distribution, multiple second ground partition confidence levels can be obtained.

[0074] The confidence calculation method for the second water area partition can be the same as the confidence calculation method for the second ground area partition. It is only necessary to replace the ratio of the number of toads and the mean of the toad number distribution with the ratio of the ventilation frequency and the mean of the ventilation frequency distribution. It will not be repeated here.

[0075] By adding the first and second confidence scores corresponding to a ground partition, we can obtain the confidence score for that ground partition (e.g., 1.667 + 2.5 = 4.167). By adding the first and second confidence scores corresponding to a water partition, we can obtain the confidence score for that water partition (this value can be 4.3, 3.8, etc.). Using the above method, we can calculate the confidence score for each ground partition and water partition, thereby obtaining the multiple confidence scores.

[0076] In the embodiment of the present application, the environmental parameter indexing unit 13 performs ideal ground humidity indexing and ideal dissolved oxygen indexing according to the gill degeneration progress distribution, and obtains the ideal ground humidity distribution and the ideal dissolved oxygen distribution, including:

[0077] Inputting the gill degradation progress corresponding to the first ground partition within the gill degradation progress distribution into a first ground humidity index table, and indexing to obtain a first ideal ground humidity, wherein the first ground humidity index table includes mapping pairs of sample ideal humidity and sample gill degradation progress of the first ground partition;

[0078] Continue to obtain multiple ideal ground humidity according to the gill degeneration progress corresponding to the multiple ground partitions, and obtain the ideal ground humidity distribution;

[0079] Inputting the gill degradation progress corresponding to the first water area partition within the gill degradation progress distribution into a first dissolved oxygen index table, and indexing to obtain a first ideal dissolved oxygen, wherein the first dissolved oxygen index table includes mapping pairs of sample ideal dissolved oxygen and sample gill degradation progress of the first water area partition;

[0080] Continue to obtain multiple ideal dissolved oxygen according to the gill degeneration progress corresponding to multiple water areas, and obtain the ideal dissolved oxygen distribution.

[0081] In an embodiment of the present application, the ideal ground humidity and ideal dissolved oxygen content in the water required for breeding toads with different gill degeneration progress can be obtained through existing toad breeding data and stored in an ideal parameter index table.

[0082] For example, when gill degeneration progresses to 80%, the ideal ground humidity can be 60%, and the ideal dissolved oxygen content can be 8 mg / L. By collecting the ideal ground humidity and ideal dissolved oxygen corresponding to the gill degeneration progress of different ground and water zones (covering 1% to 100%), an index table of ideal parameters for toad farming can be obtained.

[0083] After predicting the progress of gill degeneration through the above steps, the ideal ground humidity and ideal dissolved oxygen can be obtained from the ideal parameter index table.

[0084] Specifically, by inputting the gill degeneration progress corresponding to a first floor partition (e.g., 80%) into a first floor humidity index table, a first ideal floor humidity corresponding to the first floor partition (e.g., 60% humidity) can be indexed. By indexing multiple ideal floor humidity values ​​based on the gill degeneration progress corresponding to multiple floor partitions, an ideal floor humidity distribution can be obtained. The first floor humidity index table includes mappings of sample ideal humidity values ​​and sample gill degeneration progress values ​​for the first floor partition. For example, when the gill degeneration progress is 80%, the mapped sample ideal humidity can be 60%.

[0085] Next, the gill degeneration progress corresponding to the first water zone (e.g., 80%) is entered into the first dissolved oxygen index table to index the ideal dissolved oxygen value corresponding to that water zone (e.g., 6 mg / L). Multiple ideal dissolved oxygen values ​​are indexed based on the gill degeneration progress corresponding to multiple water zones, yielding an ideal dissolved oxygen distribution. The first dissolved oxygen index table includes mappings of ideal dissolved oxygen values ​​and gill degeneration progress values ​​for samples in the first water zone. For example, when the gill degeneration progress is 80%, the mapped ideal dissolved oxygen value can be 6 mg / L.

[0086] In the aquaculture parameter adaptation unit 14 of the embodiment of the present application, based on the Internet of Things, the actual humidity distribution of the multiple ground partitions and the actual dissolved oxygen distribution of the multiple water partitions are collected, and the aquaculture parameter adaptation degree is calculated for the ideal ground humidity distribution and the ideal dissolved oxygen distribution in combination with the multiple confidence levels to obtain the aquaculture parameter adaptation degree.

[0087] Based on the Internet of Things, actual humidity of the multiple ground zones is collected to obtain actual humidity distribution, and actual dissolved oxygen of the multiple water zones is collected to obtain actual dissolved oxygen distribution;

[0088] Calculating the deviation between the actual humidity distribution and the ideal ground humidity distribution, and calculating the degree of adaptation to obtain a humidity adaptation degree distribution;

[0089] Calculating the deviation between the actual dissolved oxygen distribution and the ideal dissolved oxygen distribution, and calculating the fitness to obtain a dissolved oxygen fitness distribution;

[0090] According to the multiple confidence levels, weighted calculation is performed on the fitness within the humidity fitness distribution and the dissolved oxygen fitness distribution to obtain the breeding parameter fitness, and toad breeding monitoring and management is performed.

[0091] The actual humidity of the ground zones can be obtained using humidity sensors deployed in each zone. For example, a temperature and humidity sensor with an IP68 protection rating (such as the HTW-211) can be used. This sensor supports a humidity range of 0-100% RH and an accuracy of ±2% RH, making it suitable for long-term use in high-humidity environments. The sensor measures the actual humidity (e.g., 80%) in each zone in real time to obtain the actual humidity distribution.

[0092] The actual dissolved oxygen level in each water zone can be measured using dissolved oxygen sensors deployed in each zone. Industrial-grade optical dissolved oxygen sensors, such as the JXBS-3001-DO, have a range of 0-20 mg / L and an accuracy of ±0.1 mg / L. The actual dissolved oxygen level (e.g., 7.0 mg / L) can be collected from multiple water zones to determine the actual dissolved oxygen distribution.

[0093] For the above-mentioned sensor system, STM32F103 or ESP32 can be used as the core controller, with an integrated ADC module to collect analog signals, and the sensor can be connected through the UART / SPI interface. The actual dissolved oxygen distribution and actual humidity distribution can be pushed to the cloud platform using the MQTT protocol to facilitate dynamic detection and adjustment of management strategies.

[0094] Furthermore, it is necessary to calculate the deviation from the ideal ground humidity distribution based on the actual humidity distribution measured by the above method, and calculate the degree of adaptation based on the deviation value. The humidity adaptation degrees corresponding to multiple ground partitions are calculated to obtain a humidity adaptation degree distribution. For example, if the actual humidity of a ground partition is 80%, and the ideal ground humidity obtained by indexing in the above steps is 60%, then the deviation between the actual humidity (80%) and the ideal ground humidity (60%) is the absolute value of the difference between the actual humidity distribution and the ideal ground humidity distribution, such as |80%-60%|=20%. The adaptation degree in this embodiment is the degree of similarity between the actual value and the ideal value, and here it is the degree of similarity between the actual humidity and the ideal ground humidity. It can be calculated by subtracting the ratio of the deviation degree to the ideal ground humidity from 1. For example, in the above example, the humidity adaptation degree = 1-(20% / 60%) ≈ 0.667.

[0095] The greater the humidity adaptability, the smaller the gap between the ideal ground humidity and the actual humidity. By calculating the humidity adaptability corresponding to each ground partition using the above method, the humidity adaptability distribution can be obtained.

[0096] Using the same acquisition method as humidity adaptability, the dissolved oxygen adaptability (such as 0.7) of a water zone can be calculated based on the actual dissolved oxygen distribution and ideal dissolved oxygen distribution of the water zone. The dissolved oxygen adaptability distribution can be calculated based on the actual dissolved oxygen distribution and ideal dissolved oxygen distribution of each water zone.

[0097] Finally, it is necessary to perform weighted calculation on the fitness within the humidity fitness distribution and the dissolved oxygen fitness distribution based on the multiple confidence levels to obtain the fitness of the breeding parameters and conduct toad breeding monitoring and management.

[0098] For example, if the confidence level for a surface zone is 4.1, and the sum of the confidence levels for all surface zones within the surface area is 41, the confidence weight for that surface zone is 4.1 / 41 = 0.1. This method can be used to calculate the confidence weight for each surface zone (or water zone). Next, multiply the confidence weight for each surface zone by the corresponding humidity suitability. For example, if the confidence weight is 0.1 and the corresponding humidity suitability is 0.667, the product is 0.0667. Adding the product of the confidence weight for each surface zone and the corresponding humidity suitability yields the aquaculture parameter suitability for the surface zone (e.g., 0.6). Similarly, multiplying the confidence weight for each water zone by the corresponding dissolved oxygen suitability and summing the results yields the aquaculture parameter suitability for the surface zone (e.g., 0.7).

[0099] Next, you can set thresholds for the adaptability of farming parameters for the ground and water areas, such as 0.5 or 0.6 (these thresholds can be adjusted based on the actual use of the system and user needs). When the adaptability of farming parameters for the ground or water area falls below the threshold (such as 0.5 or 0.6), an alarm is triggered, reminding the farming staff to manage the area and perform operations such as humidifying the ground or oxygenating the water to ensure that the toads are always in the optimal development environment.

[0100] It should be noted that, in the above embodiments, the description of each embodiment has its own focus. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0101] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0102] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0103] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0104] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0105] Although preferred embodiments of the present invention have been described, additional changes and modifications to these embodiments may occur to those skilled in the art once the basic inventive concepts become known.

[0106] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the present invention and its equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. The intelligent digital monitoring and management system for toad breeding based on the Internet of Things is characterized by: The system comprises: A toad information collection unit is used to collect the number of toads in multiple ground zones in the ground area through an infrared sensor array deployed in the ground area based on the Internet of Things, thereby obtaining the toad population distribution; and to collect the ventilation frequency of multiple water zones in the water area through an infrared sensor array deployed in the water area, thereby obtaining the ventilation frequency distribution; The gill degeneration degree prediction unit is used to predict the gill degeneration progress according to the toad population distribution and ventilation frequency distribution, and obtain the gill degeneration progress distribution and multiple confidence levels, including: Obtain the number of toads raised in the current batch, and calculate the number of toads that have not degenerated based on the distribution of the number of toads; Inputting multiple toad numbers within the toad number distribution into a ground gill degeneration prediction network, and predicting and outputting multiple ground gill degeneration progresses; inputting the non-degenerated numbers into a water area gill degeneration prediction network in combination with multiple ventilation frequencies within the ventilation frequency distribution, and predicting and outputting multiple water area gill degeneration progresses; Combining the plurality of terrestrial gill degeneration progresses and the plurality of water gill degeneration progresses to obtain a gill degeneration progress distribution; Processing to obtain a plurality of confidence levels according to the gill degeneration progress distribution; An environmental parameter indexing unit is used to perform ideal ground humidity indexing and ideal dissolved oxygen indexing according to the gill degeneration progress distribution, so as to obtain an ideal ground humidity distribution and an ideal dissolved oxygen distribution; The breeding parameter adaptation unit is used to collect the actual humidity distribution of the multiple ground partitions and the actual dissolved oxygen distribution of the multiple water partitions based on the Internet of Things, calculate the breeding parameter adaptation of the ideal ground humidity distribution and the ideal dissolved oxygen distribution in combination with the multiple confidence levels, obtain the breeding parameter adaptation, and perform toad breeding monitoring and management, including: Based on the Internet of Things, actual humidity of the multiple ground zones is collected to obtain actual humidity distribution, and actual dissolved oxygen of the multiple water zones is collected to obtain actual dissolved oxygen distribution; Calculating the deviation between the actual humidity distribution and the ideal ground humidity distribution, and calculating the degree of adaptation to obtain a humidity adaptation degree distribution; Calculating the deviation between the actual dissolved oxygen distribution and the ideal dissolved oxygen distribution, and calculating the fitness to obtain a dissolved oxygen fitness distribution; According to the multiple confidence levels, weighted calculation is performed on the fitness within the humidity fitness distribution and the dissolved oxygen fitness distribution to obtain the breeding parameter fitness, and toad breeding monitoring and management is performed.

2. The intelligent digital monitoring and management system for toad breeding based on the Internet of Things according to claim 1 is characterized in that: Based on the Internet of Things, an infrared sensor array deployed on the ground area collects the number of toads in multiple ground zones in the ground area and obtains the toad population distribution. An infrared sensor array deployed on the water surface area collects the ventilation frequency of multiple water zones in the water surface area and obtains the ventilation frequency distribution, including: Based on the Internet of Things, an infrared sensor array is deployed in the ground area to collect ground infrared sensing data of multiple ground partitions in the ground area; Based on multiple ground infrared sensor data, multiple toad populations are identified and their distribution is obtained. Based on the Internet of Things, an infrared sensor array is deployed on the water surface area to collect water area infrared sensing data of multiple water areas within a preset time period in the water surface area; Based on infrared sensing data from multiple water areas, multiple ventilation quantities are identified and obtained, and multiple ventilation frequencies are calculated to obtain ventilation frequency distribution.

3. The intelligent digital monitoring and management system for toad breeding based on the Internet of Things according to claim 1 is characterized in that: The training steps of the terrestrial gill degradation prediction network and the water gill degradation prediction network include: Based on the gill degeneration observation data of toad farming, the number of sample toads in the ground partition was collected, as well as the number of samples without degeneration and the ventilation frequency of samples in the water partition were collected; Based on the gill degeneration observation data of toad farming, the gill degeneration progress set of samples was collected and annotated; Using machine learning, we constructed a terrestrial gill degeneration prediction network and an aquatic gill degeneration prediction network; The sample gill degeneration progress set is used as output supervision data, and the sample toad quantity set is used as input training data to perform supervised training on the terrestrial gill degeneration prediction network until convergence; The sample gill degradation progress set is used as output supervision data, the sample non-degraded quantity set and the sample ventilation frequency set in the water area partition are used as input training data, and the water area gill degradation prediction network is supervised and trained until convergence.

4. The intelligent digital monitoring and management system for toad breeding based on the Internet of Things according to claim 1 is characterized in that: Based on the gill degeneration progress distribution, multiple confidence levels are obtained, including: Calculating the error margin of each gill degeneration progress and the mean of the gill degeneration progress distribution, and calculating and obtaining a plurality of first confidence levels; Calculate the ratio of each toad number to the mean of the toad number distribution to obtain multiple second ground partition confidences; Calculating the ratio of each ventilation frequency to the mean of the ventilation frequency distribution to obtain a plurality of second water area partition confidences; A plurality of confidence levels are calculated based on the first confidence level and the second confidence level of each ground partition, and the first confidence level and the second confidence level of each water partition.

5. The intelligent digital monitoring and management system for toad breeding based on the Internet of Things according to claim 1 is characterized in that: According to the gill degeneration progress distribution, an ideal ground humidity index and an ideal dissolved oxygen index are respectively performed to obtain an ideal ground humidity distribution and an ideal dissolved oxygen distribution, including: Inputting the gill degradation progress corresponding to the first ground partition within the gill degradation progress distribution into a first ground humidity index table, and indexing to obtain a first ideal ground humidity, wherein the first ground humidity index table includes mapping pairs of sample ideal humidity and sample gill degradation progress of the first ground partition; Continue to obtain multiple ideal ground humidity according to the gill degeneration progress corresponding to the multiple ground partitions, and obtain the ideal ground humidity distribution; Inputting the gill degradation progress corresponding to the first water area partition within the gill degradation progress distribution into a first dissolved oxygen index table, and indexing to obtain a first ideal dissolved oxygen, wherein the first dissolved oxygen index table includes mapping pairs of sample ideal dissolved oxygen and sample gill degradation progress of the first water area partition; Continue to obtain multiple ideal dissolved oxygen according to the gill degeneration progress corresponding to multiple water areas, and obtain the ideal dissolved oxygen distribution.

Images