Estimation system and method for simulation mathematical model parameter pairs based on big data

Abstract

The invention belongs to the technical field of industrial control simulation, and discloses an estimation system and method for simulation mathematical model parameter pairs based on big data. The system comprises a starting module, a parameter input module, a central control module, a parameter identification module, a parameter calibration module, a mathematical model building module, a simulation optimization module, a big data processing module and a display module. Through the setting of the parameter identification module, only a step signal needs to be applied to the industrial process, so that the industrial process can be operated smoothly in the system modeling process, and when the operation is stable, the acquisition of required data is finished; the algorithm of three characteristic parameters is simple, the computation amount is small, and the requirement of the computation capability of an industrial controller is not high; and at the same time, simulation data is optimized through a simulation optimization module generative adversarial network, so that the representation of the simulation data is closer to real data, the quality and accuracy of the simulation datacan be ensured, the authenticity and reliability of the simulation data can be improved to a certain extent, and the construction cost of a simulator can also be reduced.

Description

technical field [0001] The invention belongs to the technical field of industrial control simulation, and in particular relates to a system and method for estimating parameter pairs of simulation mathematical models based on big data. Background technique [0002] The design of industrial control systems relies on mathematical models of the controlled process. The dynamic characteristics of most industrial processes can be approximated by a first-order time-delay model. When the simulation mathematical model is applied to a specific industrial production device, targeted parameter estimation is required. Simulation, that is, the use of project models to transform uncertainties specific to a specific level into their impact on objectives, which is expressed at the level of project simulation as a whole. Project simulation uses computer models and a specific level of risk estimation, generally using the Monte Carlo method for simulation. Use the model to reproduce the essen...

AI Technical Summary

Problems solved by technology

However, the existing industrial simulation mathematical model modeling method requires that the applied input signal must meet the condition of "continuous excitation"; in order to obtain enough data, the system must run for a long time; in addition, the calculation of parameters is relatively large, which is difficult for engineering The application brings a lot of inconvenience; at the same time, the simulation data screening mainly depends on the premise that there are enough high-quality samples in the simulation data, and the high-quality samples have sufficient diversity, which requires strict accuracy of the simulator and is difficult to implement; adding a priori requires Manpower and material resources are required to collect object information and build 3D models, which will be endless work for complex systems

[0003] In summary, the existing problems in the existing technology are: the existing industrial simulation mathematical model modeling method requires that the applied input signal must meet the condition of "continuous excitation"; in order to obtain enough data, the system must run for a long time; In addition, the calculation of parameters is relatively large, which brings a lot of inconvenience to engineering applications; at the same time, the simulation data screening mainly depends on the premise that there are enough high-quality samples in the simulation data, and the high-quality samples have sufficient diversity, and the accuracy of the simulator is strictly required. , it is very difficult to realize; adding a priori requires manpower and material resources to collect object information and build a 3D model, and it will be endless work for complex systems

[0004] Existing systems or software use cloud servers or other servers to store data. First, data security cannot be guaranteed, and second, data integrity cannot be judged.

Images