OPTIMIPRO | Copyright 2025.
——Modeling free, breaking through the bottleneck of modeling has the ability to automatically adapt to changes in raw materials, equipment modification, etc.
In the production process, adjusting the main operating conditions online to maximize the objective function, such as yield and economic benefits, is a real-time optimization task. Traditional real-time optimization techniques are based on process modeling methods, which obtain mathematical descriptions of the optimized process through model establishment, and then use computers to calculate the optimal operating point in the current situation online.
? The optimization of production process can be divided into two types:
1、Offline optimization: Establish a mathematical model of the device and calculate optimization variables offline based on this model.
2、Real time optimization: During the production process, adjust key operating conditions (optimization variables) online to maximize optimization goals such as yield and benefits.
At present, both offline and online optimization technologies are based on the mathematical model of the device as the optimization foundation. Therefore, the success or failure of process modeling becomes the key to everything.
? The traditional process model is basically obtained through two ways：
Mechanism modeling: Using basic equations such as reaction mechanisms, thermodynamic equilibrium, and mass equilibrium in chemical engineering, establish mathematical models for each unit. Based on the process flow, integrate the models of each unit to obtain a mathematical description of the entire device.
Data modeling: Using daily operational data or arranging certain tests to obtain device testing data, and then using statistical methods to establish empirical models.
Obviously, this technology faces two inevitable problems that hinder its widespread application.
1、Difficulties in modeling.
2、The system is large and complex, unable to adapt to changes in raw materials, and has high maintenance costs in the later stage.
3、Static models can generate biases and result in unstable long-term uses.
Research abroad has shown that deficiencies in process models and optimization techniques are the main reasons for the unsatisfactory long-term benefits of enterprise IT projects. White D. C found through research that only 16% of IT projects achieved their initial budget, schedule, and functional goals. According to an analysis of approximately 250 existing commercial online optimization systems, most of them can achieve technical and commercial success in the early stages, but the long-term benefits are often not ideal.
If there is a method that can overcome the above problems, it will be an important advancement in online optimization technology. The Dynamic Correlation Integration optimization method is such a technique. After long-term theoretical research and application, many industrial applications have been achieved, resulting in long-term stable economic growth.
? A universal online optimization technology for continuous production processes
? No modeling required
? Self adaptive: able to automatically adapt to changes in raw materials, catalysts, equipment modifications, and other actual process model changes
? No additional testing required
? Strong anti-interference ability
? Strong robustness
? Low maintenance volume
? Easy to implement