A livestock and poultry breeding environment control method, device and Internet of Things system

By constructing a dynamic environmental control system of T-S fuzzy neural network and NARX/Elman neural network based on WOA optimization, the problem of insufficient accuracy of environmental regulation in livestock and poultry houses in traditional methods is solved, and high efficiency is achieved for temperature, humidity and harmful gas concentrations Control improves the robustness and fault tolerance of the system.

CN118426312BActive Publication Date: 2024-11-08GUIZHOU YUAO AGRICULTURAL TECHNOLOGY CO LTD
View PDF 5 Cites 0 Cited by

Patent Information

Application Number
CN202410514383.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-04-26
Publication Date
2024-11-08
Estimated Expiration
2044-04-26

AI Technical Summary

Technical Problem

Traditional livestock and poultry farming environment control methods are difficult to achieve precise regulation, especially the regulation of temperature, humidity and harmful gas concentrations, and the application of the Internet of Things in the livestock and poultry farming field has not yet been explored in depth.

Method used

Using WOA-optimized T-S fuzzy neural network model, NARX neural network model and Elman neural network model, combined with information processing module and multi-sensor data, a dynamic environment control system is built, and environmental parameters are realized through parallel connection between multi-layer neural network and fuzzy system. Accurate adjustment and prediction.

Benefits of technology

It improves the prediction accuracy and control accuracy of environmental parameters, enhances the robustness and fault tolerance of the system, can effectively deal with environmental noise, and achieves accurate control of temperature, humidity and harmful gas concentration.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118426312B_ABST
    Figure CN118426312B_ABST
Patent Text Reader

Abstract

The present invention discloses a livestock and poultry breeding environment control method, device and Internet of Things system, wherein the livestock and poultry breeding environment control Internet of Things system comprises a livestock and poultry house information collection terminal, a livestock and poultry house environment control terminal, a livestock and poultry breeding monitoring cloud server terminal and a livestock and poultry breeder monitoring terminal, wherein the livestock and poultry breeding monitoring cloud server terminal respectively communicates bidirectionally with the livestock and poultry house information collection terminal, the livestock and poultry house environment control terminal and the livestock and poultry breeder monitoring terminal; the livestock and poultry house information collection terminal monitors the temperature, humidity, light intensity, dust concentration, carbon dioxide concentration, hydrogen sulfide concentration, ammonia concentration and livestock and poultry house video information in the livestock and poultry house environment in real time, and uploads it to the livestock and poultry breeding monitoring cloud server terminal; the livestock and poultry house environment control terminal is used to adjust the state of the livestock and poultry house environment adjustment equipment so that the livestock and poultry house environmental parameters are suitable for livestock and poultry breeding needs; the livestock and poultry breeding monitoring cloud server terminal stores and manages the livestock and poultry environment monitoring data and provides livestock and poultry breeding data processing services for the livestock and poultry breeder monitoring terminal; the livestock and poultry breeder monitoring terminal realizes monitoring of the livestock and poultry house breeding environment and the livestock and poultry growth process. The present invention can realize the precise adjustment of livestock and poultry house breeding environment parameters.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to a livestock and poultry breeding environment control method, device and Internet of Things system. Background Art

[0002] Precise regulation of livestock and poultry breeding environmental parameters (such as temperature, humidity, and concentration of harmful gases) is one of the key factors to improve the benefits of livestock and poultry breeding and reduce breeding costs. Livestock and poultry breeding environmental parameters have the characteristics of time-varying, nonlinear, and multivariable coupling. Traditional livestock and poultry breeding environmental control methods are mostly based on simple PID control, which has the problem of large environmental regulation errors and is difficult to meet the needs of precise regulation of livestock and poultry breeding environmental parameters. On the other hand, the Internet of Things has achieved remarkable results in many fields, but its application in the field of livestock and poultry breeding is still in the exploratory stage. Summary of the invention

[0003] Purpose of the invention: The first purpose of the present invention is to provide a livestock and poultry breeding environment control method capable of realizing precise adjustment of livestock and poultry breeding environment parameters; the second purpose of the present invention is to provide a livestock and poultry breeding environment control device; the third purpose of the present invention is to provide a livestock and poultry breeding environment control Internet of Things system.

[0004] Technical solution: The livestock and poultry breeding environment control method of the present invention comprises:

[0005] Collect multiple temperature sensor outputs, multiple humidity sensor outputs and multiple harmful gas sensor outputs;

[0006] Constructing the environmental control model of livestock and poultry houses:

[0007] The livestock and poultry house environment control model includes a TS fuzzy neural network model of WOA-NARX neural network model of WOA 1-3, a livestock and poultry house environment controller 1-3 and a TS fuzzy neural network model of WOA-Elman neural network model of WOA 1-2;

[0008] The temperature setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 1 of WOA, the humidity setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 2 of WOA, and the harmful gas concentration setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 3 of WOA; the temperature control value output by TS fuzzy neural network model of WOA - NARX neural network model 1 of WOA is used as the input of livestock house environment controller 1, the humidity control value output by TS fuzzy neural network model of WOA - NARX neural network model 2 of WOA is used as the input of livestock house environment controller 2, and the harmful gas concentration control value output by TS fuzzy neural network model of WOA - NARX neural network model 3 of WOA is used as the input of livestock house environment controller 3;

[0009] The outputs of multiple temperature sensors are used as inputs of livestock and poultry house environment controller 1, the outputs of multiple humidity sensors are used as inputs of livestock and poultry house environment controller 2, and the outputs of multiple harmful gas sensors are used as inputs of livestock and poultry house environment controller 3;

[0010] The temperature detection value output by the livestock and poultry house environment controller 1 and the humidity detection value output by the livestock and poultry house environment controller 2 are respectively used as the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA, and the temperature prediction value output by the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA are respectively used as the corresponding inputs of the livestock and poultry house environment controller 1, the TS fuzzy neural network model of WOA-NARX neural network model 1 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 2 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 3 of WOA and the TS fuzzy neural network model of WOA-Elman neural network model 2 of WOA;

[0011] The humidity prediction values ​​output by the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA are respectively used as the corresponding inputs of the livestock house environment controller 2, the TS fuzzy neural network model of WOA-NARX neural network model 1 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 2 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 3 of WOA and the TS fuzzy neural network model of WOA-Elman neural network model 2 of WOA;

[0012] The harmful gas concentration detection value output by the livestock and poultry house environment controller 3 is used as the corresponding input of the TS fuzzy neural network model of WOA-the Elman neural network model 2 of WOA, and the harmful gas concentration prediction value output by the TS fuzzy neural network model of WOA-the Elman neural network model 2 of WOA is used as the corresponding input of the livestock and poultry house environment controller 3;

[0013] The output of livestock and poultry house environment controller 1 is used as the input of temperature equipment to control the temperature of the livestock and poultry house breeding environment; the output of livestock and poultry house environment controller 2 is used as the input of humidity equipment to control the humidity of the livestock and poultry house breeding environment; the output of livestock and poultry house environment controller 3 is used as the input of harmful gas purification equipment to control the concentration of harmful gases in the livestock and poultry house breeding environment.

[0014] Further, the livestock and poultry house environment controller includes a TS fuzzy neural network model of WOA-NARX neural network model of WOA, a TS fuzzy neural network model of WOA-Elman neural network model of WOA, an Elman neural network model of WOA, a NARX neural network model of WOA and an information processing module;

[0015] The parameter control value and the parameter prediction value are respectively used as the corresponding inputs of the TS fuzzy neural network model of WOA-NARX neural network model of WOA, the time series difference between the output of the TS fuzzy neural network model of WOA-NARX neural network model of WOA and the time series difference change rate of the output of the NARX neural network model of WOA and the time series difference are used as the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model of WOA, the time series value of the output of the TS fuzzy neural network model of WOA-Elman neural network model of WOA is used as the input of the Elman neural network model of WOA, the output values ​​of multiple parameter sensors are used as the input of the information processing module, the time series difference between the output of the information processing module and the output of the Elman neural network model of WOA and the time series difference change rate are respectively used as the input of the NARX neural network model of WOA and the corresponding input of the TS fuzzy neural network model of WOA-Elman neural network model of WOA, the output of the TS fuzzy neural network model of WOA-Elman neural network model of WOA is used as the output of the livestock and poultry house environment controller, and the output of the information processing module is used as the parameter detection value output by the livestock and poultry house environment controller.

[0016] Furthermore, the information processing module includes a noise reduction encoder model-WOA's Elman neural network model 1-n and a WOA's NARX neural network model, multiple parameter sensor outputs are respectively used as corresponding inputs of each noise reduction encoder model-WOA's Elman neural network model, multiple noise reduction encoder model-WOA's Elman neural network model outputs are used as corresponding inputs of the WOA's NARX neural network model, and the WOA's NARX neural network model output is the output value of the information processing module.

[0017] Furthermore, multiple temperature sensors, multiple humidity sensors and multiple harmful gas sensors are evenly arranged at multiple points in the livestock and poultry breeding environment.

[0018] The livestock and poultry breeding environment control device of the present invention comprises a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the livestock and poultry breeding environment control method is implemented.

[0019] The livestock and poultry breeding environment control Internet of Things system of the present invention comprises a livestock and poultry house information collection terminal 1-n, a livestock and poultry house environment control terminal 1-n, a livestock and poultry breeding monitoring cloud server terminal and a livestock and poultry breeder monitoring terminal 1-n, and the livestock and poultry breeding monitoring cloud server terminal communicates with the livestock and poultry house information collection terminal, the livestock and poultry house environment control terminal and the livestock and poultry breeder monitoring terminal in two directions respectively; the livestock and poultry house information collection terminal monitors the temperature, humidity, light intensity, dust concentration, carbon dioxide concentration, hydrogen sulfide concentration, ammonia concentration and livestock and poultry house video information in the livestock and poultry house environment in real time, and uploads it to the livestock and poultry breeding monitoring cloud server terminal; the livestock and poultry house environment control terminal is used to adjust the state of the livestock and poultry house environment adjustment equipment so that the livestock and poultry house environment parameters are suitable for the needs of livestock and poultry breeding; the livestock and poultry breeding monitoring cloud server terminal stores and manages the livestock and poultry environment monitoring data and provides livestock and poultry breeding data processing services for the livestock and poultry breeder monitoring terminal; the livestock and poultry breeder monitoring terminal realizes the monitoring of the livestock and poultry house breeding environment and the livestock and poultry growth process.

[0020] Furthermore, the livestock and poultry house information collection terminal includes a temperature sensor, a humidity sensor, a light intensity sensor, a dust sensor, a carbon dioxide sensor, a hydrogen sulfide sensor, an ammonia sensor, an STM32 microprocessor, a camera and a 4G communication module; each sensor is connected to the STM32 microprocessor through a corresponding conditioning circuit, and the camera and the 4G communication module are connected to the STM32 microprocessor respectively; multiple livestock and poultry house information collection terminals build a self-organizing communication network through the 4G communication module to realize information interaction with the livestock and poultry breeding monitoring cloud server.

[0021] Furthermore, the livestock and poultry house environment control terminal includes an STM32 microprocessor, and a 4G communication module, fan equipment, humidity equipment, temperature equipment, lighting equipment and harmful gas purification equipment respectively connected to the STM32 microprocessor; by configuring the livestock and poultry house environment control model in the STM32 microprocessor, automatic control of the livestock and poultry house environmental parameters is realized; the livestock and poultry house environment control terminal realizes information interaction with the livestock and poultry breeding monitoring cloud server through the 4G communication module.

[0022] Furthermore, the livestock and poultry farmer monitoring end uses an industrial control computer or mobile phone APP to realize information interaction with the livestock and poultry farming monitoring cloud server through the 5G communication network.

[0023] Furthermore, the functions of the livestock and poultry farmer monitoring terminal include livestock and poultry house communication parameter setting, livestock and poultry house data analysis, livestock and poultry house data management and early warning, livestock and poultry house data collection and control command issuance.

[0024] Beneficial effects: Compared with the prior art, the present invention has the following significant advantages:

[0025] (1) The Elman neural network model is divided into four layers: input layer, hidden layer, receiving layer and output layer. The connection between the input layer, hidden layer and output layer is similar to that of a feedforward network. The units in the input layer only transmit livestock and poultry breeding environment signals, while the units in the output layer perform the linear weighting of livestock and poultry breeding environment parameters. The transfer function of the hidden layer units can be a linear or nonlinear function. The receiving layer is also called the context layer or state layer. It is used to memorize the output value of the livestock and poultry breeding environment parameters of the hidden layer units at the previous moment and can be considered as a one-time delay operator. The characteristic of the Elman neural network model is that the output of the hidden layer is self-connected to the livestock and poultry breeding environment information input of the hidden layer through the delay and storage of the receiving layer. This self-connection method makes it sensitive to the data of the historical state of livestock and poultry breeding environment information. The addition of the internal feedback network increases the ability of the network itself to process dynamic information, thereby achieving the purpose of dynamic modeling, and strengthening the memory of the Elman neural network model for the time series feature information of livestock and poultry breeding environment information, thereby improving the prediction accuracy of the livestock and poultry breeding environment information input signal.

[0026] (2) The NARX neural network model is a dynamic recursive network model that establishes the NARX neural network model by introducing a delay module for inputting livestock and poultry breeding environment parameters and a feedback delay module. It is a sequence of multiple livestock and poultry breeding environment information time parameters that are extended along the time axis direction along the livestock and poultry breeding environment information input parameters and feedback parameters, and realizes the data correlation modeling idea of ​​function simulation function. This method establishes a model for predicting livestock and poultry breeding environment information input parameters through the input and feedback parameters of livestock and poultry breeding environment information over a period of time. The output of the NARX neural network model is used as the livestock and poultry breeding environment information input in the feedback function and is trained in a closed loop to improve the calculation accuracy of the neural network for livestock and poultry breeding environment information. The NARX neural network model realizes continuous dynamic prediction of the input information state of livestock and poultry breeding environment information.

[0027] (3) The TS fuzzy neural network model adopts a parallel connection method of neural network and fuzzy system. They share a common livestock and poultry breeding environment information input. In this way, the two parts of the network do not interfere with each other and can process livestock and poultry breeding environment information data in parallel, which improves the training speed of the network. The network also concentrates the advantages of neural networks and fuzzy reasoning systems, which enhances the robustness and fault tolerance of the network, improves the livestock and poultry breeding environment information approximation ability of the TS fuzzy neural network model, and greatly enhances the ability of the TS fuzzy neural network model to calculate livestock and poultry breeding environment information and generalize livestock and poultry breeding environment information, thereby optimizing the livestock and poultry breeding environment parameters.

[0028] (4) The temperature, humidity and harmful gas concentration of the livestock and poultry house environment have the characteristics of nonlinearity, large hysteresis, and complex dynamic changes. The sensors that measure the environmental parameters of the livestock and poultry house are easily interfered, so the measurement of the environmental parameters of the livestock and poultry house often contains large noise. In the information processing module, multiple noise reduction encoders are combined with the Elamn neural network model and the NARX neural network model. Through the compression and decompression process of the output information of the livestock and poultry house environmental temperature, humidity and harmful gas concentration sensors, the redundant information can be used to suppress its measurement noise. In the process of livestock and poultry house environmental data processing, the livestock and poultry house environmental measurement parameters are predicted and integrated, which can greatly improve the accuracy of livestock and poultry house environmental parameters.

[0029] (5) In the livestock and poultry house environment controller, the TS fuzzy neural network model of WOA-NARX neural network model of WOA optimizes the output according to the parameter prediction value, and the TS fuzzy neural network model of WOA-Elman neural network model of WOA optimizes the output according to the error and error change between the time series parameter control quantity prediction value and the parameter detection value, forming a two-level dynamic optimization parameter control quantity, thereby improving the dynamic performance and robustness of the output parameter control quantity. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 is a structural schematic diagram of the livestock and poultry house environment control model in an embodiment of the present invention;

[0031] Figure 2 is a schematic diagram of the structure of the livestock and poultry house environment controller in an embodiment of the present invention;

[0032] Figure 3 is a schematic diagram of the structure of an information processing module in an embodiment of the present invention;

[0033] Figure 4 Schematic diagram of the structure of the livestock and poultry breeding environment control Internet of Things system in an embodiment of the present invention;

[0034] Figure 5 This is a structural diagram of a livestock and poultry house information collection terminal in an embodiment of the present invention;

[0035] Figure 6 is a structural schematic diagram of the livestock and poultry house environment control terminal in an embodiment of the present invention;

[0036] Figure 7 It is a functional diagram of the livestock and poultry farmer monitoring terminal software in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described below in conjunction with the accompanying drawings.

[0038] The embodiment of the present invention provides a method for controlling livestock and poultry breeding environment, which specifically includes the following steps:

[0039] (1) Collect the outputs of multiple temperature sensors, multiple humidity sensors and multiple harmful gas sensors, and arrange multiple temperature sensors, multiple humidity sensors and multiple harmful gas sensors in a balanced manner in the livestock and poultry breeding environment. (2) Construct a livestock and poultry house environmental control model

[0040] like Figure 1 As shown, the livestock house environment control model includes WOA's TS fuzzy neural network model-WOA's NARX neural network model 1-3, livestock house environment controller 1-3 and WOA's TS fuzzy neural network model-WOA's Elman neural network model 1-2.

[0041] The temperature setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 1 of WOA, the humidity setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 2 of WOA, and the harmful gas concentration setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 3 of WOA. The temperature control value output by TS fuzzy neural network model of WOA - NARX neural network model 1 of WOA is used as the input of livestock house environment controller 1, the humidity control value output by TS fuzzy neural network model of WOA - NARX neural network model 2 of WOA is used as the input of livestock house environment controller 2, and the harmful gas concentration control value output by TS fuzzy neural network model of WOA - NARX neural network model 3 of WOA is used as the input of livestock house environment controller 3.

[0042] The outputs of multiple temperature sensors are used as inputs of the livestock house environment controller 1 , the outputs of multiple humidity sensors are used as inputs of the livestock house environment controller 2 , and the outputs of multiple harmful gas sensors are used as inputs of the livestock house environment controller 3 .

[0043] The temperature detection value output by the livestock and poultry house environment controller 1 and the humidity detection value output by the livestock and poultry house environment controller 2 are respectively used as the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA, and the temperature prediction value output by the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA are respectively used as the corresponding inputs of the livestock and poultry house environment controller 1, the TS fuzzy neural network model of WOA-NARX neural network model 1 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 2 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 3 of WOA and the TS fuzzy neural network model of WOA-Elman neural network model 2 of WOA.

[0044] The humidity prediction values ​​output by WOA's TS fuzzy neural network model-WOA's Elman neural network model 1 are respectively used as the corresponding inputs of livestock house environment controller 2, WOA's TS fuzzy neural network model-WOA's NARX neural network model 1, WOA's TS fuzzy neural network model-WOA's NARX neural network model 2, WOA's TS fuzzy neural network model-WOA's NARX neural network model 3 and WOA's TS fuzzy neural network model-WOA's Elman neural network model 2.

[0045] The harmful gas concentration detection value output by the livestock and poultry house environment controller 3 is used as the corresponding input of the TS fuzzy neural network model of WOA-the Elman neural network model 2 of WOA, and the harmful gas concentration prediction value output by the TS fuzzy neural network model of WOA-the Elman neural network model 2 of WOA is used as the corresponding input of the livestock and poultry house environment controller 3.

[0046] The output of livestock and poultry house environment controller 1 is used as the input of temperature equipment to control the temperature of the livestock and poultry house breeding environment; the output of livestock and poultry house environment controller 2 is used as the input of humidity equipment to control the humidity of the livestock and poultry house breeding environment; the output of livestock and poultry house environment controller 3 is used as the input of harmful gas purification equipment to control the concentration of harmful gases in the livestock and poultry house breeding environment.

[0047] like Figure 2 As shown, the livestock and poultry house environment controller includes a TS fuzzy neural network model of WOA-a NARX neural network model of WOA, a TS fuzzy neural network model of WOA-an Elman neural network model of WOA, an Elman neural network model of WOA, a NARX neural network model of WOA and an information processing module.

[0048] The parameter control value and the parameter prediction value are respectively used as the corresponding inputs of the TS fuzzy neural network model of WOA-NARX neural network model of WOA, the time series difference between the output of the TS fuzzy neural network model of WOA-NARX neural network model of WOA and the time series difference change rate of the output of the NARX neural network model of WOA and the time series difference are used as the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model of WOA, the time series value of the output of the TS fuzzy neural network model of WOA-Elman neural network model of WOA is used as the input of the Elman neural network model of WOA, the output values ​​of multiple parameter sensors are used as the inputs of the information processing module, the time series difference between the output of the information processing module and the output of the Elman neural network model of WOA and the time series difference change rate are respectively used as the inputs of the NARX neural network model of WOA and the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model of WOA, the output of the TS fuzzy neural network model of WOA-Elman neural network model of WOA is used as the output of the livestock and poultry house environment controller for adjusting the environment adjustment equipment. The output of the information processing module is the parameter detection value output by the livestock and poultry house environment controller.

[0049] like Figure 3As shown, the information processing module includes a noise reduction encoder model-WOA's Elman neural network model 1-n and a WOA's NARX neural network model, multiple parameter sensor outputs are respectively used as corresponding inputs of each noise reduction encoder model-WOA's Elman neural network model, multiple noise reduction encoder model-WOA's Elman neural network model outputs are used as corresponding inputs of the WOA's NARX neural network model, and the WOA's NARX neural network model output is the output value of the information processing module (detected parameter output value).

[0050] 1. Information processing module design

[0051] The information processing module includes a noise reduction encoder model - an Elman neural network model of WOA and a NARX neural network model of WOA.

[0052] 1) Design of Elman neural network model of denoising encoder model-WOA

[0053] The Elman neural network model of the denoising encoder model-WOA is a denoising encoder model connected in series with the Elman neural network model of the WOA, and the output of the denoising encoder model is used as the input of the Elman neural network model of the WOA.

[0054] The denoising autoencoder converts the high-dimensional data of livestock and poultry breeding environment parameters into low-dimensional data by training a multi-layer neural network with a small center layer. The denoising autoencoder is a typical three-layer neural network. There is a livestock and poultry breeding environment information encoding process between the hidden layer and the input layer, and a livestock and poultry breeding environment information decoding process between the output layer and the hidden layer. The autoencoder obtains the encoded representation through the encoding operation of the livestock and poultry breeding environment parameters, and obtains the reconstructed livestock and poultry breeding environment parameter decoder through the output decoding operation of the hidden layer. The data in the hidden layer is the dimensionality reduction data. Then the reconstruction error function is defined to measure the learning effect of the livestock and poultry breeding environment information autoencoder. Based on the error function, constraints can be added to generate autoencoders of various livestock and poultry breeding environment information types.

[0055] The encoder, decoder, and loss function are as follows:

[0056] Encoder: h = δ (Wx + b) (1)

[0057] Decoder:

[0058] Loss function:

[0059] Among them, h is the output value of the hidden layer; δ is the activation function, generally using the Sigmoid function or the tanh function; W and W' are weight matrices; b and b' are biases; x is the input vector; is the output vector. The training process of the denoising autoencoder is similar to that of the BP neural network.

[0060] The denoising autoencoder trains the sparse autoencoder network by adding noise data to the livestock and poultry breeding environment data. Due to the role of the livestock and poultry breeding environment noise data, the livestock and poultry breeding environment data features learned by the autoencoder network are more robust. The autoencoder network is divided into an encoding process and a decoding process. The encoding process is from the input layer to the hidden layer, and the decoding process is from the hidden layer to the output layer. The goal of the autoencoder network is to use the error function to make the input and output as close as possible. By minimizing the error function through back propagation, the optimal weights and bias of the autoencoder network are obtained to prepare for the establishment of a deep autoencoder network model. According to the encoding and decoding principle of the autoencoder network, the encoding data and the decoding data are obtained using the livestock and poultry house environment data containing noise. Finally, the error function is constructed by the decoded data and the original data. The error function is minimized by back propagation to obtain the optimal network weights and bias. The original data of the livestock and poultry house environment is destroyed by adding noise, and then the damaged data is input into the neural network as the input layer. The reconstruction result of the denoising autoencoder neural network should be similar to the original data of the livestock and poultry house environment. This method can eliminate the disturbance of the livestock and poultry house environment and obtain a stable structure. The original livestock and poultry house environment input data is obtained by adding noise to the interference input, which is then input into the encoder to obtain feature expression and then mapped to the output layer through the decoder.

[0061] The topological structure of the Elman neural network model mainly includes input layer, hidden layer, receiving layer and output layer. The main function of the input layer is to transmit the livestock and poultry breeding environment information signal; the main function of the hidden layer is to connect the feedback of the output layer and the receiving layer, and to affect the input data according to the adjustment of the weight; the receiving layer, as a one-step delay operator of the livestock and poultry breeding environment information, can make the system more sensitive to the data of the historical state of the livestock and poultry breeding environment information to achieve the purpose of memory, enhance the ability of the network itself to process the dynamic information of the livestock and poultry breeding environment to realize dynamic modeling; the main function of the output layer is linear weighted output.

[0062] The nonlinear state space expression of the Elman neural network model is:

[0063] x(k)=f(w1x c (k)+w2u(k-1))(4)

[0064] x c (k) = x(k-1)(5)

[0065] y(k)=g(w3x(k))(6)

[0066] Where k is the current moment, k-1 is the previous moment; x is the node unit vector of the hidden layer; f is the transfer function of the hidden layer neural network; w1 is the connection weight between the receiving layer and the hidden layer; w2 is the connection weight between the input layer and the hidden layer; w3 is the connection weight between the hidden layer and the output layer; x c is the feedback state vector; u is the input vector; y is the output node vector; g is the transfer function of the output neuron, which is a linear combination of the hidden layer output.

[0067] The Elman neural network model of WOA is the WOA optimized Elman neural network model. WOA is a meta-heuristic optimization algorithm that is derived from the simulation of the hunting behavior of humpback whale groups in nature, such as the whale group's search, encirclement, pursuit and attack on prey to achieve the purpose of optimization search. Compared with other optimization algorithms, WOA has the advantages of simple mechanism, few parameters and strong optimization ability. WOA optimizes the weights and thresholds of the Elman neural network model, and enhances the training speed and global optimization ability of the Elman neural network model.

[0068] The steps of WOA optimization of Elman neural network model are as follows:

[0069] Step 1: Initialize the parameters of the Elman neural network model.

[0070] Step 2: Initialize WOA parameters such as the number of whales, maximum number of iterations, and upper and lower limits of parameters.

[0071] Step 3: WOA iteratively optimizes the Elman neural network model by surrounding the prey, updating the position in a spiral manner, and searching for the prey until the iteration ends and the best individual is formed, including weights and thresholds.

[0072] Step 4: Analyze the best individual, use the analyzed weights and thresholds to train the Elman neural network model, and make predictions.

[0073] 2) Design of NARX neural network model for WOA

[0074] The NARX neural network model is a type of recurrent neural network, which consists of three parts: input layer, hidden layer and output layer. The NARX neural network model adds a time delay and feedback mechanism for livestock and poultry breeding environment information. The output value is fed back to the input value as part of the NARX structure, which enhances the memory ability of the historical data of livestock and poultry breeding environment information. It also reduces the number of livestock and poultry breeding environment information input parameters, making the model more effective.

[0075] The NARX neural network model can be expressed as:

[0076] y(t)=f[x(t-1),x(t-2),…,x(td),y(t-1),y(t-2),…,y(td)](7)

[0077] Where y(t) is the output factor, f is a nonlinear function, x(t) is the input factor, and d is the time delay. y(t) is obtained by the previous d values ​​of x(t) and the previous d values ​​of y(t) through function f.

[0078] The NARX neural network model of WOA is a WOA optimized NARX neural network model, and the design process of the WOA optimized Elman neural network model in the embodiment of the present invention can be referred to.

[0079] 2. Design of Environmental Controller for Livestock and Poultry Houses

[0080] The livestock and poultry house environment controller includes a TS fuzzy neural network model of WOA-NARX neural network model of WOA, a TS fuzzy neural network model of WOA-Elman neural network model of WOA, an Elman neural network model of WOA, a NARX neural network model of WOA and an information processing module. The design method of the Elman neural network model of WOA and the NARX neural network model of WOA refers to the design process of the information processing module in the embodiment of the present invention.

[0081] 3) TS fuzzy neural network model of WOA-NARX neural network model design of WOA

[0082] The TS fuzzy neural network model of WOA-NARX neural network model of WOA is a series connection of the TS fuzzy neural network model of WOA and the NARX neural network model of WOA, and the output of the TS fuzzy neural network model of WOA is used as the input of the NARX neural network model of WOA.

[0083] The TS fuzzy neural network model consists of a precondition network and a postcondition network. It is a nonlinear fuzzy reasoning model with the advantages of simple calculation and easy combination with adaptive methods.

[0084] Pre-Network:

[0085] The first layer is the input layer, and the number of nodes in this layer is n. The second layer is the fuzzification layer, which fuzzifies the input livestock and poultry breeding environment information data, and each neuron executes the corresponding membership function The third layer is the fuzzy rule layer. The number of nodes in the fourth layer is m, and this layer implements normalized calculation.

[0086] Post-processing network:

[0087] The first layer is the input layer, where the input value of the 0th node is x0=1, and its function is to provide the constant term of the consequent of the livestock and poultry breeding environment information fuzzy rule. The second layer has m nodes, and its function is to calculate the consequent of each rule:

[0088]

[0089] Among them, p jk is the connection weight between layer 1 and layer 2, x k It is the input of TS fuzzy neural network model.

[0090] Layer 3 computes system output:

[0091]

[0092] Among them, y is the output of TS fuzzy neural network model, is the output of the predecessor network, y j is the output of the post-network. The central value c of the membership function of the second layer is adjusted through the network learning algorithm j and width b j and the connection weight p of the subsequent network jk For simplicity, the parameter p jk The consequent of each rule is fixed, and the subsequent of each rule becomes the connection weight of one layer in the simplified structure. This simplified structure has the same structure as the fuzzy neural network of the conventional model, and the calculation results of the conventional model can be applied.

[0093] The TS fuzzy neural network model of WOA is the process of WOA optimizing the TS fuzzy neural network model, and the NARX neural network model of WOA is the process of WOA optimizing the NARX neural network model. Refer to the WOA optimized Elman neural network model design process and related design methods in the embodiments of the present invention.

[0094] 4) TS fuzzy neural network model of WOA-Elman neural network model design of WOA

[0095] The TS fuzzy neural network model of WOA-Elman neural network model of WOA is a series connection of the TS fuzzy neural network model of WOA and the Elman neural network model of WOA, and the output of the TS fuzzy neural network model of WOA is the input of the Elman neural network model of WOA. The design method of the TS fuzzy neural network model of WOA and the Elman neural network model of WOA refers to the relevant design process of the embodiment of the present invention.

[0096] 3. Design of Environmental Control Model for Livestock and Poultry Houses

[0097] The livestock house environment control model includes the TS fuzzy neural network model of WOA-the NARX neural network model of WOA, the livestock house environment controller and the TS fuzzy neural network model of WOA-the Elman neural network model of WOA. The design process of the TS fuzzy neural network model of WOA-the NARX neural network model of WOA, the livestock house environment controller and the TS fuzzy neural network model of WOA-the Elman neural network model of WOA refers to the relevant design method of the embodiment of the present invention.

[0098] The embodiment of the present invention also provides a livestock and poultry breeding environment control device, comprising a memory and a processor, wherein a computer program is stored in the memory, and when the computer program is executed by the processor, the livestock and poultry breeding environment control method described in the embodiment of the present invention is implemented. The livestock and poultry breeding environment control device has all the advantages of the livestock and poultry breeding environment control method.

[0099] like Figure 4 As shown, the embodiment of the present invention also provides an Internet of Things system for controlling the environment of livestock and poultry breeding, including livestock and poultry house information collection terminals 1-n, livestock and poultry house environment control terminals 1-n, livestock and poultry breeding monitoring cloud server terminals and livestock and poultry breeder monitoring terminals 1-n, and the livestock and poultry breeding monitoring cloud server terminals communicate bidirectionally with the livestock and poultry house information collection terminals, the livestock and poultry house environment control terminals and the livestock and poultry breeder monitoring terminals respectively. The livestock and poultry house information collection terminal monitors the temperature, humidity, light intensity, dust concentration, carbon dioxide concentration, hydrogen sulfide concentration, ammonia concentration and livestock and poultry house video information in the livestock and poultry house environment in real time, and uploads them to the livestock and poultry breeding monitoring cloud server terminal. The livestock and poultry house environment control terminal is used to adjust the state of the livestock and poultry house environment adjustment equipment so that the livestock and poultry house environmental parameters are suitable for livestock and poultry breeding needs. The livestock and poultry breeding monitoring cloud server terminal stores and manages the livestock and poultry environment monitoring data and provides livestock and poultry breeding data processing services for the livestock and poultry breeder monitoring terminal. The livestock and poultry breeder monitoring terminal realizes monitoring of the livestock and poultry house breeding environment and the livestock and poultry growth process.

[0100] like Figure 5 As shown, the livestock and poultry house information collection terminal includes a temperature sensor, a humidity sensor, an illumination sensor, a dust sensor, a carbon dioxide sensor, a hydrogen sulfide sensor, an ammonia sensor, an STM32 microprocessor, a camera and a 4G communication module. Each sensor is connected to the STM32 microprocessor through a corresponding conditioning circuit, and the camera and the 4G communication module are connected to the STM32 microprocessor. Multiple livestock and poultry house information collection terminals build a self-organizing communication network through the 4G communication module to realize information interaction with the livestock and poultry breeding monitoring cloud server.

[0101] The software of the livestock and poultry house information collection terminal mainly realizes the communication of 4G communication module, the collection of livestock and poultry house environmental parameters and livestock and poultry house video information. The software adopts C language programming, with high compatibility, which greatly improves the work efficiency of software design and development, and enhances the reliability, readability and portability of program code.

[0102] like Figure 6 As shown, the livestock and poultry house environment control terminal includes an STM32 microprocessor, and a 4G communication module, a fan device, a humidity device, a temperature device, a lighting device, and a harmful gas purification device connected to the STM32 microprocessor respectively. By configuring the above-mentioned livestock and poultry house environment control model in the STM32 microprocessor, automatic control of the livestock and poultry house environmental parameters is achieved. The livestock and poultry house environment control terminal realizes information interaction with the livestock and poultry breeding monitoring cloud server through the 4G communication module.

[0103] The livestock and poultry farmer monitoring terminal uses an industrial control computer or mobile phone APP. The livestock and poultry farmer monitoring terminal mainly monitors livestock and poultry environmental parameters and realizes information exchange with the livestock and poultry farming monitoring cloud server through the 5G communication network. Figure 7 As shown in the figure, the main functions of the livestock and poultry farmer monitoring terminal include livestock and poultry house communication parameter setting, livestock and poultry house data analysis, livestock and poultry house data management and early warning, livestock and poultry house data collection and control command issuance. In addition, the livestock and poultry farmer monitoring terminal also provides functions such as livestock and poultry house breeding parameter monitoring, livestock and poultry house historical data query, personnel management and breeding logs.

Claims

1. A livestock and poultry breeding environment control method, characterized in that: include: Collect multiple temperature sensor outputs, multiple humidity sensor outputs and multiple harmful gas sensor outputs; Constructing the environmental control model of livestock and poultry houses: The livestock and poultry house environment control model includes a TS fuzzy neural network model of WOA-NARX neural network model of WOA 1-3, a livestock and poultry house environment controller 1-3 and a TS fuzzy neural network model of WOA-Elman neural network model of WOA 1-2; The temperature setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 1 of WOA, the humidity setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 2 of WOA, and the harmful gas concentration setting value is used as the input of TS fuzzy neural network model of WOA - NARX neural network model 3 of WOA; the temperature control value output by TS fuzzy neural network model of WOA - NARX neural network model 1 of WOA is used as the input of livestock house environment controller 1, the humidity control value output by TS fuzzy neural network model of WOA - NARX neural network model 2 of WOA is used as the input of livestock house environment controller 2, and the harmful gas concentration control value output by TS fuzzy neural network model of WOA - NARX neural network model 3 of WOA is used as the input of livestock house environment controller 3; The outputs of multiple temperature sensors are used as inputs of livestock and poultry house environment controller 1, the outputs of multiple humidity sensors are used as inputs of livestock and poultry house environment controller 2, and the outputs of multiple harmful gas sensors are used as inputs of livestock and poultry house environment controller 3; The temperature detection value output by the livestock and poultry house environment controller 1 and the humidity detection value output by the livestock and poultry house environment controller 2 are respectively used as the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA, and the temperature prediction value output by the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA are respectively used as the corresponding inputs of the livestock and poultry house environment controller 1, the TS fuzzy neural network model of WOA-NARX neural network model 1 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 2 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 3 of WOA and the TS fuzzy neural network model of WOA-Elman neural network model 2 of WOA; The humidity prediction values ​​output by the TS fuzzy neural network model of WOA-Elman neural network model 1 of WOA are respectively used as the corresponding inputs of the livestock house environment controller 2, the TS fuzzy neural network model of WOA-NARX neural network model 1 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 2 of WOA, the TS fuzzy neural network model of WOA-NARX neural network model 3 of WOA and the TS fuzzy neural network model of WOA-Elman neural network model 2 of WOA; The harmful gas concentration detection value output by the livestock and poultry house environment controller 3 is used as the corresponding input of the TS fuzzy neural network model of WOA-the Elman neural network model 2 of WOA, and the harmful gas concentration prediction value output by the TS fuzzy neural network model of WOA-the Elman neural network model 2 of WOA is used as the corresponding input of the livestock and poultry house environment controller 3; The output of livestock and poultry house environment controller 1 is used as the input of the temperature device to control the temperature of the livestock and poultry house breeding environment; the output of livestock and poultry house environment controller 2 is used as the input of the humidity device to control the humidity of the livestock and poultry house breeding environment; the output of livestock and poultry house environment controller 3 is used as the input of the harmful gas purification device to control the concentration of harmful gases in the livestock and poultry house breeding environment; The livestock and poultry house environment controller includes a TS fuzzy neural network model of WOA-NARX neural network model of WOA, a TS fuzzy neural network model of WOA-Elman neural network model of WOA, an Elman neural network model of WOA, a NARX neural network model of WOA and an information processing module; The parameter control value and the parameter prediction value are respectively used as the corresponding inputs of the TS fuzzy neural network model of WOA-NARX neural network model of WOA, the time series difference between the output of the TS fuzzy neural network model of WOA-NARX neural network model of WOA and the time series difference change rate of the output of the NARX neural network model of WOA and the time series difference are used as the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model of WOA, the time series value of the output of the TS fuzzy neural network model of WOA-Elman neural network model of WOA is used as the input of the Elman neural network model of WOA, the output values ​​of multiple parameter sensors are used as the inputs of the information processing module, the time series difference between the output of the information processing module and the output of the Elman neural network model of WOA and the time series difference change rate are respectively used as the inputs of the NARX neural network model of WOA and the corresponding inputs of the TS fuzzy neural network model of WOA-Elman neural network model of WOA, the output of the TS fuzzy neural network model of WOA-Elman neural network model of WOA is used as the output of the livestock and poultry house environment controller, and the output of the information processing module is used as the parameter detection value output by the livestock and poultry house environment controller; The information processing module includes Elman neural network models 1-n of the noise reduction encoder model-WOA and the NARX neural network model of the WOA. The outputs of multiple parameter sensors are respectively used as the corresponding inputs of each Elman neural network model of the noise reduction encoder model-WOA, and the outputs of the Elman neural network models of multiple noise reduction encoder models-WOA are used as the corresponding inputs of the NARX neural network model of the WOA. The output of the NARX neural network model of the WOA is the output value of the information processing module.

2. The livestock and poultry breeding environment control method according to claim 1, characterized in that: Multiple temperature sensors, multiple humidity sensors and multiple harmful gas sensors are evenly arranged at multiple points in the livestock and poultry breeding environment.

3. A livestock and poultry breeding environment control device, comprising a memory and a processor, wherein a computer program is stored in the memory, characterized in that: When the computer program is executed by a processor, the livestock and poultry breeding environment control method according to claim 1 or 2 is implemented.

4. A livestock and poultry breeding environment control Internet of Things system, which implements the livestock and poultry breeding environment control method according to claim 1 or 2, characterized in that: It includes livestock and poultry house information collection terminals 1-n, livestock and poultry house environment control terminals 1-n, livestock and poultry breeding monitoring cloud server terminals and livestock and poultry breeder monitoring terminals 1-n. The livestock and poultry breeding monitoring cloud server terminals communicate with the livestock and poultry house information collection terminals, livestock and poultry house environment control terminals and livestock and poultry breeder monitoring terminals in two directions respectively; the livestock and poultry house information collection terminals monitor the temperature, humidity, light intensity, dust concentration, carbon dioxide concentration, hydrogen sulfide concentration, ammonia concentration and livestock and poultry house video information in the livestock and poultry house environment in real time, and upload them to the livestock and poultry breeding monitoring cloud server terminals; the livestock and poultry house environment control terminals are used to adjust the status of the livestock and poultry house environment adjustment equipment to make the livestock and poultry house environment parameters suitable for livestock and poultry breeding needs; the livestock and poultry breeding monitoring cloud server terminals store and manage livestock and poultry environment monitoring data and provide livestock and poultry breeding data processing services for the livestock and poultry breeder monitoring terminals; the livestock and poultry breeder monitoring terminals monitor the livestock and poultry house breeding environment and the livestock and poultry growth process.

5. The livestock and poultry breeding environment control Internet of Things system according to claim 4, characterized in that: The livestock and poultry house information collection terminal includes temperature sensors, humidity sensors, light intensity sensors, dust sensors, carbon dioxide sensors, hydrogen sulfide sensors, ammonia sensors, STM32 microprocessors, cameras and 4G communication modules; each sensor is connected to the STM32 microprocessor through the corresponding conditioning circuit, and the camera and 4G communication module are connected to the STM32 microprocessor respectively; multiple livestock and poultry house information collection terminals build a self-organizing communication network through the 4G communication module to realize information interaction with the livestock and poultry breeding monitoring cloud server.

6. The livestock and poultry breeding environment control Internet of Things system according to claim 5, characterized in that: The livestock and poultry house environment control terminal includes an STM32 microprocessor, and a 4G communication module, fan equipment, humidity equipment, temperature equipment, lighting equipment and harmful gas purification equipment respectively connected to the STM32 microprocessor; by configuring the livestock and poultry house environment control model in the STM32 microprocessor, automatic control of the livestock and poultry house environmental parameters can be achieved; the livestock and poultry house environment control terminal realizes information interaction with the livestock and poultry breeding monitoring cloud server through the 4G communication module.

7. The livestock and poultry breeding environment control Internet of Things system according to claim 6, characterized in that: The livestock and poultry farmer monitoring terminal uses an industrial control computer or mobile phone APP to achieve information exchange with the livestock and poultry farming monitoring cloud server through the 5G communication network.

8. The livestock and poultry breeding environment control Internet of Things system according to claim 7, characterized in that: The functions of the livestock and poultry farmer monitoring terminal include livestock and poultry house communication parameter setting, livestock and poultry house data analysis, livestock and poultry house data management and early warning, livestock and poultry house data collection and control command issuance.

Citation Information

Patent Citations

  • Decoupling fuzzy control method of closed livestock and poultry breeding environment

    CN107203239A

  • Intelligent purification system for harmful gas in livestock and poultry house environment

    CN113031555A

  • Livestock and poultry house breeding environment parameter intelligent control system

    CN113126676A

  • Greenhouse inspection robot system and multi-source information fusion data processing method

    CN117784611A

  • Breeding environment temperature prediction and control system for livestock and poultry house, and adjustment and control method therefor

    WO2022166919A1