Complex industrial process data-driven whole plant level performance optimization control method

Abstract

The invention relates to an industrial process optimization control method, in particular to a complex industrial process data-driven whole plant level performance optimization control method. The method comprises the following process of multi-objective multi-constraint whole plant level performance optimization control which comprises a, construction of an objective function, wherein in order toachieve whole plant level production indexes, an ideal value is tracked in an approximate optimal mode to serve as a target, and the objective function is composed of tracking errors and a quadraticform of operation indexes; b, description of constraint conditions, wherein production index dynamics, the limited production indexes, the limited operation indexes and limited interference input areadopted as the constraint conditions; c, multi-objective multi-constraint optimization control. The invention also includes an optimal operation index design based on a non-zero-sum game, an estimation algorithm design of an optimal operation index design based on ADP and simulation and physical experiment verification; a technical route combining theoretical analysis, simulation experiments and laboratory physical experiments is adopted, so that the method not only has a strong theoretical and scientific meaning but also maintains high practicability.

Description

technical field [0001] The invention relates to an industrial data optimization control method, in particular to a complex industrial process data-driven whole plant level performance optimization control method. Background technique [0002] At present, research on the design of complex industrial applications has attracted widespread attention, especially the problems that need to be solved in the aspect of operation optimization control include: [0003] a) The dynamics of production indicators are nonlinear, and there are system dynamics that cannot be accurately modeled, and there are unknown bounded disturbances; [0004] b) How to solve multi-objective performance optimization problems. [0005] In this part of the design, there are problems such as inaccurate modeling of plant-level production index dynamics in complex industrial processes, multi-production index and multi-unit process coupling, and multi-time scale problems. The optimal operating index of the anal...

AI Technical Summary

Problems solved by technology

[0005] In this part of the design, there are complex industrial process plant-level production index dynamics that cannot be accurately modeled, multi-production index and multi-unit process coupling, and multi-time scale problems

The optimal operating index of the analysis design guarantees the conditions for realizing Nash equilibrium, and the condition of zero steady-state tracking error for the production index. The optimal operating index designed requires the dynamic information of the production index to be known, but for complex industrial processes, the plant-level production index Dynamics are difficult to accurately model

Images