Page 48 - 《橡塑技术与装备》英文版2026年3期

HINA R&P TECHNOLOGY AND EQUIPMENT

and effective implementation strategies, an efficient and system operation status, conduct anomaly detection and
environmentally friendly waste plastic pyrolysis system can be preventive maintenance, and provide decision support.
constructed, providing strong technical support for solving the 3.1.1.3 Implementation strategy
plastic pollution problem. Implement in stages: Divide the project into stages such
Refer to existing successful cases both domestically and as requirement analysis, system design, prototype development,
internationally, such as practices that utilize AI to optimize testing and verification, and online deployment, and advance
the pyrolysis process, thereby increasing yield and reducing step by step. Through a feedback loop, continuously collect
pollutant emissions. Learn from their successful experiences system operation data, optimize AI models and system
and technical details. parameters, and enhance overall performance. Ensure that the
The following is a study exploring the architecture design system design complies with environmental standards, adheres
of the AI waste plastic pyrolysis system from the perspectives to data protection laws and regulations, and safeguards the
of system requirements analysis, technology selection, safety of operators.
architecture design, and implementation strategy: Interdisciplinary Collaboration: Assemble a team of
3.1.1.1 System requirement analysis experts from fields such as chemistry, mechanics, electronics,
Conduct an in-depth analysis of the requirements for the and computer science to ensure the comprehensiveness and
entire system to be developed, including but not limited to: innovativeness of the system.
the raw data that the system needs to receive, the results that Continuous optimization: Through a feedback loop,
the system should provide, performance indicators such as continuously collect system operation data, optimize AI models
processing speed, accuracy, and stability, as well as cost and and system parameters, and enhance overall performance.
resource constraints, including budget, power consumption, Compliance and safety: Ensure that the system design
and equipment space. complies with environmental standards, adheres to data
3.1.1.2 Technical selection protection laws and regulations, and safeguards the safety of
Data collection: The front-end data collection layer is operators.
responsible for collecting and preprocessing raw data. Utilizing 3.1.2 Exploration of the design of functional
big data technologies and cloud computing platforms for data modules (types)
processing, such as Hadoop and Spark, to handle and store Each module is responsible for a specific function,
vast amounts of real-time and historical data, including data reducing complexity. Modules should be open to the outside
cleaning and feature extraction. world, allowing for expansion and modification; internally,
Model construction: Selecting appropriate AI they should be closed, not affected by external changes.
technologies, such as deep learning, reinforcement learning, or Components within a module are closely related, while
machine learning algorithms, for predicting and optimizing the connections between modules should be minimized to
pyrolysis process. reduce mutual dependencies. Utilizing mature frameworks
Control and execution: Integrate intelligent controllers and libraries, such as Spring Boot and React.js, can enhance
and actuators, such as PID controllers, to adjust the parameters development efficiency. Databases are selected based on data
of the pyrolysis process based on the output of the AI model. types and requirements, including relational databases like
AI model layer: Deploy AI models for prediction and MySQL and NoSQL databases like MongoDB.
optimization, such as using deep neural networks to predict the The functional module system should include the
yield and quality of pyrolysis products. following core modules:
Control execution layer: Based on the output of the AI 3.1.2.1 Data acquisition module
model, adjust the parameters of the pyrolysis process, such as Responsible for collecting various parameter data in
temperature and pressure, through an intelligent controller. real-time during the pyrolysis process, including temperature,
Backend monitoring and decision-making layer: Monitor pressure, reaction rate, etc., to provide a foundation for

·2· Vol.52，No.3