Dynamic yak weighing method and device based on GAN-BPNN
By applying GAN-BPNN model and sensor technology in the yak weighing device, dynamic, accurate and automated measurement of yak weight is achieved, and the problems of inefficient efficiency, large errors and insufficient applicability in the existing technology are solved.
Patent Information
- Application Number
- CN202411464162.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-21
- Publication Date
- 2025-02-07
AI Technical Summary
The prior art has problems of inefficiency, large errors and insufficient applicability when weighing yaks. Especially for yaks with large plateaus, manual fixation is required, resulting in inaccurate yak stress response and measurement.
Using the GAN-BPNN-based dynamic weighing method and device, the yak position is detected by infrared sensors, the pressure sensor collects data, and the GAN-BPNN model is used to accurately calculate the weight information to achieve fully automated dynamic weighing.
It improves the accuracy and efficiency of yak weight measurement, reduces stress response and error caused by manual intervention, realizes fully automated data collection, and reduces labor costs.
Smart Images
Figure CN119394412A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of yak weighing, and in particular to a GAN-BPNN-based yak dynamic weighing method and device. Background Art
[0002] In the beef cattle farming industry, weighing yaks is the key to implementing precise nutritional regulation, improving production efficiency and economic benefits. Traditional yak weighing methods usually use manual operations, which have problems such as low efficiency and large errors, and are difficult to meet the needs of large-scale farming. In addition, manual weighing requires a lot of time and manpower, which increases farming costs.
[0003] In view of the shortcomings of traditional manual weighing, a device for collecting cattle weight on pastures has been designed. It adopts a semi-automatic form, and obtains relevant static data by driving yaks into the weighing channel and fixing their positions. Specifically, when the cattle are driven into the weighing channel, the cattle stand completely on the load-bearing plate of the weighing platform, and the load-bearing plate is subjected to force to drive the force transmission rod to press down the weighing sensor to collect the cattle weight data. At the same time, the cattle are clamped by the relative movement of two clamping plates to limit the cattle's range of movement, so as to achieve the effect of quickly positioning and weighing the cattle. In addition, several methods for weighing and calculating yaks have been studied in the prior art: (1) Dynamic weighing is treated as a parameter estimation and prediction problem based on the least squares method, and an adaptive least squares method based on Householder transformation is used. While improving the weighing speed, the measurement accuracy of the system is also guaranteed. (2) By analyzing the movement mode of dairy cows (walking-pause, steady walking, and intense exercise), the sensor data is used to judge the state of dairy cows, and data is collected during steady walking, thereby realizing the dynamic weight measurement of dairy cows. (3) The LSTM neural network is used to predict the weight of yaks. (4) The EMD empirical mode decomposition algorithm can be used to better measure the weight of beef cattle.
[0004] Although the above research has achieved certain results, the existing devices and methods are not suitable for the large and wild yaks on the plateau. Specifically, because yaks are unique to the plateau, they are large in size and have strong wild behaviors. The above device needs to fix the yaks when weighing. For yaks with bad tempers, they will have a stress response to human intervention, which will affect the normal growth of yaks. In addition, the above device is a semi-automatic design and still requires human participation. The efficiency is low, and the collected data is not digitally stored, which makes it difficult to meet the development needs of modern breeding. In addition, if the yak exercises violently on the weighing platform, it will cause errors in the collection of pressure data, resulting in inaccurate measurements. Summary of the invention
[0005] Multiple aspects of the present application provide a GAN-BPNN-based yak dynamic weighing method and device, which measure the weight of the yak without restricting the yak's movements and improve the accuracy of the measurement.
[0006] In order to achieve the above technical effects, one aspect of the present application provides a dynamic weighing method for yaks based on GAN-BPNN, comprising:
[0007] When the yak passes through the weighing platform, the position of the yak on the weighing platform is detected;
[0008] When the yak is detected in the middle of the weighing platform, the pressure data of the yak is collected;
[0009] The pressure data is input into the GAN-BPNN model to obtain the weight information of the yak.
[0010] According to a preferred embodiment of the present invention, when the yak passes through the weighing platform, detecting the position of the yak on the weighing platform further comprises:
[0011] Infrared sensors are respectively arranged at the head end, the middle part and the end part of the weighing platform, and the position of the yak is detected by the infrared sensors;
[0012] The yak ear tag information is obtained through the wireless radio frequency sensor to determine the corresponding yak identity information.
[0013] According to a preferred embodiment of the present invention, when the yak is detected in the middle of the weighing platform, collecting the pressure data of the yak further includes:
[0014] When the infrared sensors arranged at the head and the end of the weighing platform are not triggered, and the infrared sensor arranged in the middle of the weighing platform is triggered, the pressure sensor arranged at the bottom of the weighing platform is started to collect the pressure data of the yak.
[0015] According to a preferred embodiment of the present invention, the step of inputting the pressure data into a GAN-BPNN model to obtain the weight information of the yak further comprises:
[0016] Based on historical yak weight data, a GAN adversarial network model was trained and generated;
[0017] Input random noise into the trained GAN adversarial network model to generate Fake_data simulated data;
[0018] Using the fake data to simulate the data, the BPNN neural network model is trained to obtain a GAN-BPNN model;
[0019] The pressure data of the yak to be tested is input into the GAN-BPNN model to obtain the weight information of the yak.
[0020] According to a preferred embodiment of the present invention, the training and generating of the GAN adversarial network model based on the historical yak weight data further comprises:
[0021] Input the historical yak pressure data and actual weight data into the GAN adversarial network model to be trained, and set the generator and discriminator;
[0022] Use BCEWithLogitsLoss() as the loss function;
[0023] The generator and discriminator are optimized alternately according to the loss function, and finally the Nash equilibrium is reached to generate the GAN adversarial network model.
[0024] According to a preferred embodiment of the present invention, the method of using the fake_data simulated data to train the BPNN neural network model to obtain the GAN-BPNN model further comprises:
[0025] The fake_data simulation data is used as training data, and the historical yak pressure data and actual weight data are used as test sets to input into the BPNN neural network model to be trained;
[0026] The mean square gradient descent function is selected as the loss function;
[0027] The model parameters were adjusted according to the error between the yak weight data output by the model and the actual weight data to generate the GAN-BPNN model.
[0028] According to a preferred embodiment of the present invention, the method further comprises:
[0029] The model training samples are updated regularly, and the GAN-BPNN model parameters are updated using the updated training samples.
[0030] Another aspect of the present application provides a yak dynamic weighing device based on GAN-BPNN, comprising:
[0031] Position detection module: used to detect the position of the yak on the weighing platform when the yak passes through the weighing platform;
[0032] Data acquisition module: used to collect the pressure data of the yak when the yak is detected in the middle of the weighing platform;
[0033] Weight measurement module: used to input the pressure data into the GAN-BPNN model to obtain the weight information of the yak.
[0034] According to a preferred embodiment of the present invention, the device also includes a weighing platform, which is an inverted trapezoidal structure, with infrared sensors arranged on both sides connected to the position detection module, and a pressure sensor arranged at the bottom of the weighing platform connected to the data acquisition module.
[0035] Another aspect of the present application provides a yak dynamic weighing device based on GAN-BPNN, the device comprising:
[0036] at least one processor; and
[0037] a memory communicatively connected to the at least one processor; wherein,
[0038] The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method described above.
[0039] In another aspect of the present application, a computer-readable storage medium is provided, on which computer program instructions are stored. The computer program instructions can be executed by a processor to implement the above method.
[0040] In the solution provided in the embodiment of the present application, the Internet of Things is combined with artificial intelligence, infrared sensors are used to determine the position of the yak, and the GAN-BPNN model is used to accurately calculate the yak's weight, thereby reducing the error caused by the dynamic walking of the yak, avoiding the problem of the need to fix the yak for weighing, reducing the yak's stress response, making the measurement results more accurate, and fully automated collection, which reduces labor costs and improves the efficiency of yak weight collection. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, a brief introduction will be given below to the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative labor.
[0042] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:
[0043] Figure 1 A schematic diagram of a dynamic weighing method for yaks based on GAN-BPNN provided in one embodiment of the present application;
[0044] Figure 2 A computational flow chart of a GAN generative adversarial network provided in an embodiment of the application;
[0045] Figure 3 A calculation flow chart of a BPNN neural network provided for an embodiment of the application;
[0046] Figure 4A schematic diagram of the structure of a yak dynamic weighing device based on GAN-BPNN provided in one embodiment of the present application;
[0047] Figure 5 A structural schematic diagram of a weighing platform provided in one embodiment of the present application;
[0048] Figure 6 The present invention is a schematic diagram of the structure of a device suitable for implementing the solution in the embodiment of the present application.
[0049] The same or similar reference numerals in the drawings represent the same or similar components. DETAILED DESCRIPTION
[0050] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.
[0051] In a typical configuration of the present application, the terminal and the equipment of the service network each include one or more processors (CPU), input / output interface, network interface and memory.
[0052] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.
[0053] Computer readable media include permanent and non-permanent, removable and non-removable media, and can be implemented by any method or technology to store information. Information can be computer program instructions, data structures, modules of programs or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, read-only compact disk (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device.
[0054] In actual scenarios, the execution subject of the method can be a user device, or a device formed by integrating a user device and a network device through a network, or an application running on the above device. The user device includes but is not limited to various terminal devices such as computers, mobile phones, tablet computers, smart watches, and bracelets. The network device includes but is not limited to network hosts, single network servers, multiple network server sets, or cloud computing-based computer sets, which can be used to implement some processing functions when setting an alarm. Here, the cloud is composed of a large number of hosts or network servers based on cloud computing, where cloud computing is a type of distributed computing, a virtual computer composed of a group of loosely coupled computer sets.
[0055] The main cattle raised in the northwestern plateau are yaks. Yaks are large in size, wild and have a violent temper. If there is too much human intervention when weighing yaks, it will cause stress reactions such as strenuous exercise on the scale, and the collected pressure data will have large errors.
[0056] Figure 1 A schematic diagram of a dynamic weighing method for yaks based on GAN-BPNN is provided in an embodiment of the present application, as shown in FIG. Figure 1 As shown, the method includes:
[0057] S101. When the yak passes through the weighing platform, the position of the yak on the weighing platform is detected.
[0058] Specifically, a pressure sensor for detecting the weight of the yak is arranged on the weighing platform, and infrared sensors are arranged at the head end, the middle part and the tail end of the weighing platform. The infrared sensors can be through-beam infrared sensors. The through-beam infrared sensors of the present invention are arranged along the direction in which the yak walks on the weighing platform during weighing. When the yak is at different positions on the weighing platform, different through-beam infrared sensors will be triggered. Therefore, the specific position of the yak during weighing can be known according to the state of the through-beam infrared sensors, and then it can be judged whether the current position is suitable for weighing. The specific judgment method is shown in the following table:
[0059] Table 1 Weighing platform status determination table
[0060]
[0061] Infrared A is set at the head end of the weighing platform, infrared B is set in the middle of the weighing platform, and infrared C is set at the end of the weighing platform. It can be seen from Table 1 that when the three pairs of infrared sensors are all 0, the current state is considered to be empty, and it is determined that the yak has not passed the weighing platform. The state flag is 0, and the pressure sensor is calibrated at this time. When the state of the middle infrared sensor, that is, infrared B, is 1 and the infrared A and infrared C at the head and tail are in state 0, it is considered to be a normal state, and it is determined that the yak is on the weighing platform at this time. The state flag is 1, and the pressure sensor data is collected at this time. When only the infrared A state is 1, it is determined that the yak is about to enter the weighing platform, and the state flag is 2. It belongs to the normal weighing process but no data is recorded. When only the infrared C state is 1, it is determined that the yak is leaving the weighing platform, and this weighing is over. The state flag is 3, and the data is sent out through the serial port. When infrared A and infrared C are both in state 1, it is determined that there are different yaks before and after the current weighing platform, and a warning reminder is issued.
[0062] An ear tag is set on the head of each yak, and wireless radio frequency sensors are set on both sides of the weighing platform. The yak information is obtained through wireless connection with the ear tag to match it with the weight data collected subsequently.
[0063] S102: When it is detected that the yak is in the middle of the weighing platform, pressure data of the yak is collected.
[0064] Specifically, when the yak is in the middle of the weighing platform, it means that the yak is in the most stable state. At this time, the data is relatively accurate. At this time, the infrared sensors located at the head and tail ends of the weighing platform are not triggered, and the infrared sensor located in the middle of the weighing platform is triggered. After the controller receives the feedback signal from the infrared sensor, it turns on the pressure sensor to collect the yak's weight information.
[0065] S103, inputting the pressure data into the GAN-BPNN model to obtain the weight information of the yak.
[0066] Specifically, firstly, the GAN-BPNN model is trained and generated based on the historical yak weight data. The GAN-BPNN model is a combination of the GAN generative adversarial network and the BP neural network.
[0067] The core idea of GAN generative adversarial network comes from the Nash equilibrium of game theory. The two parties in the game include the generator and the discriminator. The trained GAN network can generate similar fake data according to the characteristic distribution of the input sample. The use of GAN network can effectively enrich the original yak weight dataset, reduce the requirements for the training dataset, and improve the applicability of the algorithm. It has strong practical value in practical applications. The calculation process of GAN generative adversarial network is as follows: Figure 2 shown.
[0068] BP neural network is a multi-feedforward neural network based on the back propagation error algorithm. BP neural network consists of input layer, hidden layer and output layer. Information enters the network from the input layer, passes through the hidden layer, and finally reaches the output layer. During the training process, BP neural network continuously adjusts the connection weights to make the network output as close to the expected output as possible, thereby achieving learning. Fusion prediction of weight data through BP neural network can effectively reduce the interference caused by yak movement, thereby improving measurement accuracy. The calculation process of BP neural network is as follows: Figure 3 shown.
[0069] Specifically, in the embodiment of the present invention, the GAN generative adversarial network is first trained, and the data (ADC value) collected by the pressure sensor corresponding to the historical yak is used as the weight data. The hidden layer of the generator and the discriminator is set to 2 layers. The output layer, hidden layer, and output layer node structure of the generator G are set to (2, 16, 64, 2). The output layer, hidden layer, and output layer node structure of the discriminator G are set to (2, 16, 64, 1)
[0070] Define input noise: Generate random noise as the input of the generator using a random number seed to ensure that the generator can produce diverse outputs.
[0071] Train the GAN generative adversarial network, and set the learning rate of the generator and discriminator to 0.01. Select BCEWithLogitsLoss() (binary cross entropy loss with logits) as the loss function. According to the loss function, the generator and discriminator are alternately optimized, and finally the Nash equilibrium is reached. At this time, it is considered that the GAN model has good performance, and the GAN model is determined. Then use the trained GAN generative adversarial network to input random noise and generate 3000 fake_data data to expand the original weight data.
[0072] Then, the Fake_data simulated data is used to train the BPNN neural network model to obtain the GAN-BPNN model. The training process is as follows:
[0073] 1. Dataset division:
[0074] Use the generated 3000 fake_data data as the training data of the BPNN model, and the original weight data as the test set.
[0075] 2. Data normalization processing:
[0076]
[0077] Where:
[0078] adcmax and ADC min is the maximum and minimum ADC value in the data set.
[0079] According to formula (1), all data are normalized to eliminate the influence of feature dimension, accelerate model convergence speed and improve model stability.
[0080] 3. BPNN model construction
[0081] (1) Determine the input and output: The input feature of the BPNN neural network is the given ADC value, and the output feature is the weight data of the yak.
[0082] (2) Network structure design: The number of hidden layers is set to 1, and the node structure of the input layer, hidden layer, and output layer is set to (1, 16, 1).
[0083] (3) Select the activation function structure: the activation function of the hidden layer is ReLU (Rectified Linear Unit), and the activation function of the output layer is Sigmoid.
[0084] 4. Training BPNN model:
[0085] The mean squared gradient descent (MSGD) function is selected as the loss function, Adam is used as the optimizer, and the learning rate is set to 0.01. Forward propagation stage: (1) The input data is weighted and nonlinearly transformed, and the result is passed to the hidden layer. (2) The data in the hidden layer is also weighted and nonlinearly transformed, and the result is passed to the output layer. (3) The error e is calculated based on the predicted output W of the BP neural network and the expected output W2.
[0086]
[0087] Back propagation phase:
[0088] (4) Update the weights of each part of the network based on the calculated error e.
[0089] w ij =w ij +ηH j (1-H j )x(i)w j e (3)
[0090] w j =w j +ηH j e (4)
[0091] (5) Update the network node threshold α,d according to the error e.
[0092] a j =a j +ηH j (1-H j ) j e (5)
[0093] d i =d i +e (6)
[0094] Where: W ij ——The connection weight between node i in layer L1 and node j in layer L2;
[0095] W j ——The connection weight from L2 layer j to the output layer node;
[0096] W——prediction output;
[0097] e——the error between the predicted output and the expected output;
[0098] α——hidden layer threshold;
[0099] d——output layer threshold;
[0100] η——learning rate.
[0101] When the Loss value of the model converges and approaches 0, it is considered that the model has good fitting performance and the GAN-BPNN model is determined.
[0102] Finally, the trained GAN-BPNN model is used for data fitting. The dynamic ADC value of the yak to be tested is collected by the pressure sensor and passed into the determined GAN-BPNN model to output the corresponding weight information of the yak.
[0103] Preferably, in order to ensure the accuracy of the GAN-BPNN model, the model training samples are updated regularly, and the GAN-BPNN model parameters are updated using the updated training samples.
[0104] In the method provided in the embodiment of the present invention, the Internet of Things is combined with artificial intelligence, an infrared sensor is used to determine the position of the yak, and the GAN-BPNN model is used to accurately calculate the weight of the yak, thereby reducing the error caused by the dynamic walking of the yak, avoiding the problem of the need to fix the yak for weighing, reducing the stress response of the yak, making the measurement results more accurate, and fully automated collection, which reduces labor costs and improves the efficiency of yak weight collection.
[0105] Figure 4 A schematic diagram of a dynamic weighing device for yaks based on GAN-BPNN is provided in an embodiment of the present application, as shown in FIG. Figure 4 As shown, the device comprises:
[0106] Position detection module 11: used to detect the position of the yak on the weighing platform when the yak passes through the weighing platform;
[0107] Data collection module 22: used to collect pressure data of the yak when the yak is detected in the middle of the weighing platform;
[0108] The weight measurement module 33 is used to input the pressure data into the GAN-BPNN model to obtain the weight information of the yak.
[0109] The above device can execute the poultry breeding index evaluation method in the above-mentioned embodiment, wherein the position detection module 11 executes step S101; the data acquisition module 22 executes step S102; and the weight measurement module 33 executes step S103.
[0110] Figure 5 A structural diagram of a weighing platform provided in one embodiment of the present application is shown in FIG. Figure 5 As shown, the yak dynamic weighing device also includes a weighing platform, which is an inverted trapezoidal structure. Pairs of infrared sensors are respectively arranged at the positions A, B, and C on both sides, which are connected to the position detection module 11. A pressure sensor is arranged at the bottom of the weighing platform, which is connected to the data acquisition module 22. A wireless radio frequency device RFID is placed at the front end of point A to read the yak ear tag information and match it with the subsequent collected weight data. When the yak walks steadily across the weighing platform, the position of the yak on the weighing platform is judged by the different states of the infrared beam, so as to judge whether the weight information (ADC value) of the yak is obtained. The ADC value is passed into the GAN-BPNN model of the weight measurement module 33 to obtain the yak weight data.
[0111] Based on the same inventive concept, an electronic device is also provided in an embodiment of the present application, and the method corresponding to the electronic device may be the dynamic weighing method of yaks based on GAN-BPNN in the aforementioned embodiment, and its principle of solving the problem is similar to that of the method. The electronic device provided in an embodiment of the present application includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the methods and / or technical solutions of the aforementioned multiple embodiments of the present application.
[0112] The electronic device may be a user device, or a device formed by integrating a user device and a network device through a network, or may be an application running on the above device. The user device includes but is not limited to various terminal devices such as computers, mobile phones, tablet computers, smart watches, and bracelets. The network device includes but is not limited to network hosts, single network servers, multiple network server sets, or cloud computing-based computer sets, which can be used to implement some processing functions when setting an alarm. Here, the cloud is composed of a large number of hosts or network servers based on cloud computing, where cloud computing is a type of distributed computing, a virtual computer composed of a group of loosely coupled computer sets.
[0113] Figure 6 The structure of a device suitable for implementing the method and / or technical solution in the embodiment of the present application is shown, and the device 1200 includes a central processing unit (CPU, Central Processing Unit) 1201, which can perform various appropriate actions and processes according to the program stored in the read-only memory (ROM, Read Only Memory) 1202 or the program loaded from the storage part 1208 to the random access memory (RAM, Random Access Memory) 1203. In RAM1203, various programs and data required for system operation are also stored. CPU 1201, ROM 1202 and RAM 1203 are connected to each other through bus 1204. Input / output (I / O, Input / Output) interface 1205 is also connected to bus 1204.
[0114] The following components are connected to the I / O interface 1205: an input section 1206 including a keyboard, a mouse, a touch screen, a microphone, an infrared sensor, etc.; an output section 1207 including a cathode ray tube (CRT), a liquid crystal display (LCD), an LED display, an OLED display, etc., and a speaker, etc.; a storage section 1208 including one or more computer-readable media such as a hard disk, an optical disk, a magnetic disk, a semiconductor memory, etc.; and a communication section 1209 including a network interface card such as a LAN (Local Area Network) card, a modem, etc. The communication section 1209 performs communication processing via a network such as the Internet.
[0115] In particular, the methods and / or embodiments in the embodiments of the present application may be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program includes program code for executing the method shown in the flowchart. When the computer program is executed by the central processing unit (CPU) 1201, the above functions defined in the method of the present application are executed.
[0116] Another embodiment of the present application further provides a computer-readable storage medium having computer program instructions stored thereon, wherein the computer program instructions can be executed by a processor to implement the methods and / or technical solutions of any one or more embodiments of the present application described above.
[0117] Specifically, the present embodiment may adopt any combination of one or more computer-readable media. The computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, device, or device.
[0118] Computer readable signal media may include data signals propagated in baseband or as part of a carrier wave, which carry computer readable program code. Such propagated data signals may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. Computer readable signal media may also be any computer readable medium other than a computer readable storage medium, which may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device.
[0119] The program code embodied on the computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
[0120] Computer program code for performing the operation of the present application can be written in one or more programming languages or a combination thereof, including object-oriented programming languages such as Java, Smalltalk, C++, and conventional procedural programming languages such as "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as an independent software package, partially on the user's computer and partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer can be connected to the user's computer through any type of network including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computer (e.g., using an Internet service provider to connect through the Internet).
[0121] The flow chart or block diagram in the accompanying drawings shows the possible architecture, function and operation of the equipment, method and computer program product according to various embodiments of the present application. In this regard, each square box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some implementations as replacements, the functions marked in the square box can also occur in a sequence different from that marked in the accompanying drawings. For example, two square boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each square box in the block diagram and / or flow chart, and the combination of the square boxes in the block diagram and / or flow chart can be implemented with a dedicated system for hardware that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.
[0122] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0123] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or page components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.
[0124] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
[0125] The above-mentioned integrated unit implemented in the form of a software functional unit can be stored in a computer-readable storage medium. The above-mentioned software functional unit is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform some steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), disk or optical disk and other media that can store program codes.
[0126] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.
[0127] In addition, it is clear that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. Multiple units or devices stated in a device claim can also be implemented by one unit or device through software or hardware. The words first, second, etc. are used to indicate names, and do not indicate any particular order.
Claims
1. A dynamic weighing method for yaks based on GAN-BPNN, characterized in that: include: When the yak passes through the weighing platform, the position of the yak on the weighing platform is detected; When the yak is detected in the middle of the weighing platform, the pressure data of the yak is collected; The pressure data is input into the GAN-BPNN model to obtain the weight information of the yak.
2. The dynamic weighing method of yak based on GAN-BPNN according to claim 1 is characterized in that: When the yak passes through the weighing platform, detecting the position of the yak on the weighing platform further includes: Infrared sensors are respectively arranged at the head end, the middle part and the end part of the weighing platform, and the position of the yak is detected by the infrared sensors; The yak ear tag information is obtained through the wireless radio frequency sensor to determine the corresponding yak identity information.
3. The dynamic weighing method of yak based on GAN-BPNN according to claim 2 is characterized in that: When the yak is detected in the middle of the weighing platform, collecting the pressure data of the yak further includes: When the infrared sensors arranged at the head and the end of the weighing platform are not triggered, and the infrared sensor arranged in the middle of the weighing platform is triggered, the pressure sensor arranged at the bottom of the weighing platform is started to collect the pressure data of the yak.
4. The dynamic weighing method of yak based on GAN-BPNN according to claim 1, characterized in that: The step of inputting the pressure data into the GAN-BPNN model to obtain the weight information of the yak further includes: Based on historical yak weight data, a GAN adversarial network model was trained and generated; Input random noise into the trained GAN adversarial network model to generate Fake_data simulated data; Using the fake data to simulate the data, the BPNN neural network model is trained to obtain a GAN-BPNN model; The pressure data of the yak to be tested is input into the GAN-BPNN model to obtain the weight information of the yak.
5. The dynamic weighing method of yak based on GAN-BPNN according to claim 4 is characterized in that: The training and generating of the GAN adversarial network model according to the historical yak weight data further includes: Input the historical yak pressure data and actual weight data into the GAN adversarial network model to be trained, and set the generator and discriminator; Use BCEWithLogitsLoss() as the loss function; The generator and discriminator are optimized alternately according to the loss function, and finally the Nash equilibrium is reached to generate the GAN adversarial network model.
6. The dynamic weighing method of yak based on GAN-BPNN according to claim 4, characterized in that: The method of using the fake_data simulated data to train the BPNN neural network model to obtain the GAN-BPNN model further includes: The fake_data simulation data is used as training data, and the historical yak pressure data and actual weight data are used as test sets to input into the BPNN neural network model to be trained; The mean square gradient descent function is selected as the loss function; The model parameters were adjusted according to the error between the yak weight data output by the model and the actual weight data to generate the GAN-BPNN model.
7. The dynamic weighing method of yak based on GAN-BPNN according to claim 1, characterized in that: The method further comprises: The model training samples are updated regularly, and the GAN-BPNN model parameters are updated using the updated training samples.
8. A dynamic weighing device for yaks based on GAN-BPNN, characterized in that: include: Position detection module: used to detect the position of the yak on the weighing platform when the yak passes through the weighing platform; Data acquisition module: used to collect the pressure data of the yak when the yak is detected in the middle of the weighing platform; Weight measurement module: used to input the pressure data into the GAN-BPNN model to obtain the weight information of the yak.
9. The yak dynamic weighing device based on GAN-BPNN according to claim 8, characterized in that: The device also includes a weighing platform, which is an inverted trapezoidal structure. Infrared sensors are arranged on both sides and connected to the position detection module. A pressure sensor is arranged at the bottom of the weighing platform and connected to the data acquisition module.
10. A yak dynamic weighing device based on GAN-BPNN, the device comprising: at least one processor; as well as a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to any one of claims 1 to 7.