Seed traceability system based on combination of block chain, Internet of Things and neural network

Abstract

The invention relates to the field of seed whole-process traceability, and particularly discloses a seed traceability system based on a block chain, an internet of things and a neural network. The system comprises a hardware part and a software part, the hardware part comprises a printer based on a Beidou positioning system and a portable device based on the Internet of Things, and the software part comprises a seed planting data center and a block chain traceability platform which are built on a ubuntu server of Tencent cloud. According to the method, the problems of unsafe and opaque information of the seeds from production to circulation are solved, the transparency and credibility of the whole process are ensured, the trust of consumers on seed products is enhanced, powerful support is provided for industry supervision, and the whole seed industry is promoted to develop in a safer, more efficient and sustainable direction.

Description

Technical Field

[0001] The present invention relates to the technical field of whole-process traceability of seeds, and specifically to a seed traceability system based on the combination of blockchain, Internet of Things and neural network. Background Art

[0002] At present, most of the traceability of agricultural products on the market adopts traditional database traceability, and there are problems of insecurity and opacity in the traceability data. In recent years, some traceability systems have used blockchain for traceability, but these systems do not have the support of hardware, and when entering information, it is only subjectively entered by people. It only achieves data security and transparency in the data traceability stage, and does not achieve data security and transparency in the data collection and entry stage.

[0003] The management and traceability of crop seed information greatly affect the supervision of seed quality. At present, most traceability systems only target agricultural products, and there are very few related to seeds. Secondly, most traceability systems only achieve the traceability of agricultural products from the information aspect. Many information needs to be subjectively entered by people, lacking corresponding hardware support, and it is difficult to accurately judge the authenticity of the entered information. Summary of the Invention

[0004] In view of the above problems of the prior art, the present invention provides a seed traceability system based on the combination of blockchain, Internet of Things and neural network, which solves the problems of insecurity and opacity of information in the production and circulation of seeds, and can ensure the security and transparency of the entered and traced information.

[0005] The technical solution adopted by the present invention is as follows: A seed traceability system based on the combination of blockchain, Internet of Things and neural network, including a hardware part and a software part. The hardware part includes a printer based on the Beidou positioning system and an Internet of Things portable device, and the software part includes a seed planting data center and a blockchain traceability platform built on an ubuntu server in Tencent Cloud.

[0006] Preferably, the Internet of Things portable device in the hardware part and the printer based on the Beidou positioning system use the stm32f103rct6 single-chip microcomputer as the main control, and the battery supplies power to the load end through a voltage stabilization module; the load end includes soil temperature and humidity, air temperature and humidity, and light intensity sensors, a 4G module, a printer module supporting secondary development, a Beidou positioning module, and a main control single-chip microcomputer; the neural network uses an lstm neural network, and the hyperparameters of the neural network are found using the Bayesian optimization algorithm, and an attention mechanism is added to the model.

[0007] Preferably, the printer based on the Beidou positioning system is provided with a traceability QR code, and stipulates that the user must be within the geographical range allowed by the program to print the QR code, and the geographical location information setting in the program is based on the manufacturer information provided by the user and will be verified by specialized personnel.

[0008] Preferably, the specific process of printing the traceability QR code by the printer based on the Beidou positioning system is: Step 1: After the power is turned on, the Beidou positioning module locates the current information and transmits the longitude and latitude to the main control microcontroller. The microcontroller determines whether the current longitude and latitude are within the geographical range provided by the user, and controls the power supply of the printer by closing the control relay; Step 2: Within the specified geographical location, the relay is closed, the printer is powered on normally, and starts to print; otherwise, the relay is not closed and the printer cannot be powered on.

[0009] Preferably, the portable device based on the Internet of Things is a data collecting device, and the data collecting device is composed of a temperature and humidity sensor, a soil temperature and humidity sensor, a light sensor and a camera; The process of data collection by the portable device based on the Internet of Things is as follows: Step 1: After the main control microcontroller exchanges data with each sensor, it sends the data collected by the sensor to the supporting software through the 4G module, and the data collected by the hardware can be seen in real time on the software; Step 2: The video stream collected by the camera is transmitted to the software through the EZVIZ Cloud module to obtain the status of the equipment deployment site in real time.

[0010] Preferably, the construction of the seed planting data center includes the following functions: displaying the geographic location submitted by the user, used in conjunction with a Beidou positioning printer; transmitting data collected by sensors of portable IoT devices to the seed planting data center; and transmitting video streams collected by cameras of portable IoT devices to the seed planting center.

[0011] Preferably, the specific process of building a seed planting data center to build a training neural network model is: Step 1: After setting up the IoT transmission protocol service on the Ubuntu system, debug the 4G module on the IoT portable device. The device sends data to the seed planting data center, and the seed planting data center stores the received data in the local database. Step 2: Before training the neural network model, manually change the data collected by the IoT device and record it; Step 3: Download the collected data in the seed planting data center, and label which data was collected after artificial modification and which was collected before artificial modification. Input the labeled data into the neural network model for training, and judge the authenticity of the data collected by the device. The neural network model uses the LSTM model architecture for training, finds its optimal hyperparameters through the Bayesian optimization algorithm, and adds an attention mechanism to the model.

[0012] Preferably, the blockchain used by the blockchain traceability platform is a currency-free blockchain based on Fabric, which is divided into a blockchain part and a platform part.

[0013] Preferably, the specific steps for traceability of the blockchain traceability platform are as follows: Step 1: Modify the structure and chain code part in the blockchain part, add 7 user types including breeding companies, breeding companies, processing companies, transportation companies, sales companies, consumers, and administrators. Different user types have different forms for input. After input, deploy the blockchain and start the front and back ends in the platform part. Step 2: Upload the locally trained neural network model and deploy it to two ports of the server. When entering information at the blockchain front end, the front end transmits the form data entered by the "breeding company" and "processing company" user types to the corresponding neural network ports. After the neural network makes a judgment, it transmits the judgment label back to the platform front end. After the front end obtains the judgment label, it saves the form data and the judgment label in the blockchain at the same time.

[0014] Preferably, the steps of the neural network trained for the "processing company" user type are as follows: Step 1: The seed processing enterprise provides the physical and chemical indicators of real seed processing and makes them into an excel dataset format. Step 2: Based on the real dataset format, use the random function to randomly construct the same number of false data. Step 3: Mark the real degree labels for the real data and the false data to complete the production of the training set. Step 4: Input the training set into the LSTM neural network and use the Bayesian optimization algorithm to find the optimal hyperparameters. Step 5: Use the optimal hyperparameters to train the model, save the model, and upload and deploy it to the neural network flask of the server. After inputting the physical and chemical indicators of seed processing, the real degree label is output after being judged by the neural network and then uploaded to the chain.

[0015] The present invention has the following characteristics and advantages: (1)The present invention provides a seed traceability system based on the combination of blockchain, Internet of Things and neural network, which solves the problems of insecure and opaque information in the process of seed production and circulation. (2)As the core of the system, blockchain technology ensures the security and immutability of data. Corresponding data records are generated at each link of breeding, propagation, processing, transportation and sales of each seed, and these records are securely stored and managed through blockchain technology. Due to the distributed nature of blockchain, all participants can access and verify these data in real time, thus eliminating the risks of information silos and data tampering, and ensuring the transparency and credibility of the entire supply chain. (3)The application of Internet of Things technology enables real-time monitoring of seeds at each link. By embedding sensors, the environmental conditions (such as temperature, humidity, light, etc.) of seeds during growth, transportation and storage can be collected in real time. (4)The combination of neural network provides strong support for data analysis. By performing deep learning on a large amount of collected data, the neural network can identify potential quality problems and risk factors.

[0016] The technical solution of the present invention will be further described in detail below through the accompanying drawings and embodiments. Description of the Drawings

[0017] Figure 1 is the system composition diagram of a seed traceability system based on the combination of blockchain, Internet of Things and neural network of the present invention; Figure 2 is the brief wiring diagram of a printer based on the Beidou positioning system of a seed traceability system based on the combination of blockchain, Internet of Things and neural network of the present invention; Figure 3 is the brief wiring diagram of a portable device based on the Internet of Things of a seed traceability system based on the combination of blockchain, Internet of Things and neural network of the present invention; Figure 4 is the software operation architecture diagram of a seed traceability system based on the combination of blockchain, Internet of Things and neural network of the present invention. Detailed Embodiments

[0018] The following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the present invention claimed, but merely represents selected embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0019] Such as Figure 1As shown in the figure, a seed traceability system based on the combination of blockchain, Internet of Things, and neural network provided by the present invention includes a hardware part and a software part. The hardware part includes a printer based on the Beidou positioning system and an Internet of Things-based portable device. The software part includes a seed planting data center and a blockchain traceability platform built on an ubuntu server in Tencent Cloud.

[0020] Hardware part: 1. Printer based on the Beidou positioning system Currently, some agricultural product packages are marked with QR codes, and consumers can query the traceability information of agricultural products by scanning the codes. The present invention designs a printer based on the Beidou positioning system, as Figure 2 shown. The user must be within the geographical range specified by the program to print the traceability QR code, which reduces the possibility of QR code forgery geographically. The geographical location information in the program is set based on the manufacturer information provided by the user and will be verified by a special person.

[0021] For the Internet of Things portable device and the printer based on the Beidou positioning system of the technical solution, the stm32f103rct6 single-chip microcomputer is used as the main control, and the battery powers the load end through a voltage stabilizing module. The load end includes soil temperature and humidity, air temperature and humidity, and light intensity sensors, a 4G module, a printer module supporting secondary development, a Beidou positioning module, and the main control single-chip microcomputer.

[0022] The specific process of using the printer based on the Beidou positioning system to print the traceability QR code is as follows: Step 1: After the power is turned on, the Beidou positioning module locates the current information, transmits the longitude and latitude to the main control single-chip microcomputer, and the single-chip microcomputer determines whether the current longitude and latitude are within the geographical range provided by the user, and controls the power supply of the printer by controlling the closing of the relay; Step 2: Within the specified geographical location, the relay closes, the printer is normally powered on, and starts to work for printing; otherwise, the relay does not close and the printer cannot be powered on.

[0023] 2. Internet of Things-based portable device As Figure 3 shown, the present invention designs an Internet of Things-based portable device. The devices for collecting data include: temperature and humidity sensors, soil temperature and humidity sensors, light sensors, and cameras. The stm32f103rct6 single-chip microcomputer is used as the main control, and the battery powers the load end.

[0024] After the master microcontroller conducts data inquiries with each sensor, it sends the data collected by the sensors to the supporting software through the 4G module, and the data collected by the hardware can be viewed in real time on the software. The video stream collected by the camera is transmitted to the software through the Ezviz Cloud module, and the situation of the equipment deployment site can be viewed in real time.

[0025] Software part: 1. Seed Planting Data Center Build a seed planting data center on the ubuntu server of Tencent Cloud, including but not limited to three functions: display the geographical locations submitted and approved by users, and use it in conjunction with the Beidou positioning printer; transmit the data collected by the sensors of the Internet of Things portable device to the seed planting data center; transmit the video stream collected by the camera of the Internet of Things portable device to the seed planting center.

[0026] The specific process of building and training a neural network model in the seed planting data center is as follows: Step 1: After building the Internet of Things transmission protocol service on the ubuntu system, debug the 4G module on the Internet of Things portable device. The device sends data to the seed planting data center, and the seed planting data center stores the received data in the local database; Step 2: Before building the training neural network model, artificially modify the data collected by the Internet of Things devices and record them; Step 3: Download the collected data in the seed planting data center, and mark which data was collected after artificial modification and which was collected before artificial modification. Input the labeled data into the neural network model for training, and judge the authenticity of the data collected by the device. The neural network model uses the lstm model architecture, finds its optimal hyperparameters through the Bayesian optimization algorithm, and adds an attention mechanism to the model.

[0027] 2. Blockchain Traceability Platform Build a blockchain traceability platform on the ubuntu server of Tencent Cloud. The blockchain used is a non-currency-based blockchain based on fabric, which is divided into the blockchain part and the platform part.

[0028] The specific steps for traceability of the blockchain traceability platform are as follows: Step 1: Modify the structure and chain code parts in the blockchain part, add 7 user types including breeding companies, breeding companies, processing companies, transportation companies, sales companies, consumers, and administrators. Different user types have different forms for entry. After entry, deploy the blockchain and start the front and back ends in the platform part; Step 2: Upload the locally trained neural network model and deploy it to two ports of the server. When entering information at the blockchain front-end, the front-end transmits the form data entered for user types of "breeding company" and "processing company" to the ports of the neural network. After the neural network finishes judgment, it transmits the judgment label back to the platform front-end, and the obtained judgment label and the entered form data are simultaneously saved in the blockchain.

[0029] Among them, the steps of the neural network trained for the user type of "processing company" are as follows: Step 1: The seed processing enterprise provides the physical and chemical indexes of real seed processing and makes them into the format of an excel data set. Step 2: On the basis of the real data set format, use the random function to randomly construct the same amount of false data. Step 3: Mark the real data and false data with authenticity labels to complete the production of the training set. Step 4: Input the training set into the lstm neural network and use the Bayesian optimization algorithm to find the optimal hyperparameters. Step 5: Use the optimal hyperparameters to train the model, save the model and upload it to be deployed on the neural network flask of the server. After inputting the physical and chemical indexes of seed processing, the authenticity label is output after being judged by the neural network and then uploaded to the chain.

[0030] As Figure 4 shown, the software operation of a seed traceability system provided by the present invention based on the combination of blockchain, Internet of Things and neural network is as follows: The Internet of Things portable device collects and uploads data to the seed planting center, and the seed planting center shares the data to the local database. The local database trains the model and deploys it to the neural network flask, and the neural network flask judges the information authenticity.

[0031] Users interact with the front-end, the front-end enters information, and the information with verified authenticity is uploaded to the back-end after verification, and the back-end transmits the information to the consensus mechanism, and the consensus mechanism stores the information in the blockchain.

[0032] Therefore, a seed traceability system proposed by the present invention based on the combination of blockchain, Internet of Things and neural network solves the problems of insecure and opaque information in the process of seed production to circulation, ensures the transparency and credibility of the whole process, enhances consumers' trust in seed products, and also provides strong support for industry supervision, promoting the entire seed industry to develop in a safer, more efficient and sustainable direction. Ultimately, this system will contribute to the sustainable development of global agriculture, ensuring food safety and the protection of the ecological environment.

[0033] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions of the present invention or make equivalent replacements, and these modifications or equivalent replacements cannot make the modified technical solutions deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A seed traceability system based on blockchain, Internet of Things and neural network, characterized in that: It includes a hardware part and a software part. The hardware part includes a printer based on the Beidou positioning system and a portable device based on the Internet of Things. The software part includes a seed planting data center and a blockchain traceability platform built on the Ubuntu server of Tencent Cloud.

2. According to claim 1, a seed traceability system based on the combination of blockchain, Internet of Things and neural network is characterized in that: The hardware part of the Internet of Things portable device and the printer based on the Beidou positioning system use the stm32f103rct6 microcontroller as the main control, and the battery powers the load end through the voltage stabilizing module; the load end includes soil temperature and humidity, air temperature and humidity and light intensity sensors, a 4G module, a printer module that supports secondary development, a Beidou positioning module and a main control microcontroller; the neural network uses the LSTM neural network, the hyperparameters of the neural network are found using the Bayesian optimization algorithm, and an attention mechanism is added to the model.

3. According to claim 1, a seed traceability system based on the combination of blockchain, Internet of Things and neural network is characterized in that: The printer based on the Beidou positioning system is provided with a traceability QR code, and stipulates that the user must be within the geographical range allowed by the program to print the QR code, and the geographical location information setting in the program is based on the manufacturer information provided by the user and will be verified by specialized personnel.

4. According to claim 1, a seed traceability system based on the combination of blockchain, Internet of Things and neural network is characterized in that: The specific process of printing the traceability QR code by the printer based on the Beidou positioning system is as follows: Step 1: After the power is turned on, the Beidou positioning module locates the current information and transmits the longitude and latitude to the main control microcontroller. The microcontroller determines whether the current longitude and latitude are within the geographical range provided by the user, and controls the power supply of the printer by closing the control relay; Step 2: Within the specified geographical location, the relay is closed, the printer is powered on normally, and starts to print; Otherwise, the relay is not closed and the printer cannot be powered on.

5. According to claim 1, a seed traceability system based on the combination of blockchain, Internet of Things and neural network is characterized in that: The portable device based on the Internet of Things is a device for collecting data, and the device for collecting data is composed of a temperature and humidity sensor, a soil temperature and humidity sensor, a light sensor, and a camera; The process of data collection by the portable device based on the Internet of Things is as follows: Step 1: After the main control microcontroller exchanges data with each sensor, it sends the data collected by the sensor to the supporting software through the 4G module, and the data collected by the hardware can be seen in real time on the software; Step 2: The video stream collected by the camera is transmitted to the software through the EZVIZ Cloud module to obtain the status of the equipment deployment site in real time.

6. According to claim 1, a seed traceability system based on the combination of blockchain, Internet of Things and neural network is characterized in that: The seed planting data center includes the following functions: displaying the geographic location submitted by the user, used in conjunction with the Beidou positioning printer; transmitting the data collected by the sensors of the portable IoT devices to the seed planting data center; transmitting the video stream collected by the cameras of the portable IoT devices to the seed planting center.

7. According to claim 1, a seed traceability system based on the combination of blockchain, Internet of Things and neural network is characterized in that: The specific process of building a seed planting data center to build and train a neural network model is as follows: Step 1: After setting up the IoT transmission protocol service on the Ubuntu system, debug the 4G module on the IoT portable device. The device sends data to the seed planting data center, and the seed planting data center stores the received data in the local database. Step 2: Before training the neural network model, manually change the data collected by the IoT device and record it; Step 3: Download the collected data in the seed planting data center and mark which data was collected after human modification and which was collected before human modification. Input the labeled data into the neural network model for training and judge the authenticity of the data collected by the device. The neural network model training uses the LSTM model architecture, finds its optimal hyperparameters through the Bayesian optimization algorithm, and adds an attention mechanism to the model.

8. The seed traceability system based on the combination of blockchain, Internet of Things and neural network according to claim 1 is characterized in that: The blockchain used by the blockchain traceability platform is a currency-free blockchain based on fabric, which is divided into a blockchain part and a platform part.

9. The seed traceability system based on the combination of blockchain, Internet of Things and neural network according to claim 1 is characterized in that: The specific steps of tracing the blockchain traceability platform are as follows: Step 1: Modify the structure and chain code in the blockchain part, add seven user types: breeding company, breeding company, processing company, transportation company, sales company, consumer and administrator, and enter different forms for different user types. After entering, deploy the blockchain and start the front-end and back-end in the platform part; Step 2: Upload the locally trained neural network model and deploy it to the two ports of the server. When entering information on the blockchain front end, the front end will transmit the form data entered by the user types of "breeding company" and "processing company" to the port of the corresponding neural network. After the neural network completes the judgment, it will transmit the judgment label back to the platform front end. After the front end obtains the judgment label, it will save the form data and the judgment label in the blockchain at the same time.

10. A seed traceability system based on blockchain, Internet of Things and neural network according to claim 9, characterized in that: The steps for training the neural network for the "Processing Company" user type are: Step 1: Seed processing enterprises provide real physical and chemical indicators of seed processing and make them into Excel data set format; Step 2: Based on the real data set format, use the random function to randomly construct the same amount of false data; Step 3: Label the real data and fake data with authenticity labels to complete the training set production; Step 4: Input the training set into the LSTM neural network and use the Bayesian optimization algorithm to find the optimal hyperparameters; Step 5: Use the optimal hyperparameters to train the model and save and upload the model to the neural network flask on the server. After inputting the physical and chemical indicators of seed processing, the neural network outputs the authenticity label after judgment and then uploads it to the chain.

Images