Brayton cycle system modeling method based on nonlinear programming

Abstract

The invention discloses a Brayton cycle system modeling method based on nonlinear programming. The Brayton cycle system modeling method comprises the five steps of boundary condition determination, process division, parameter explicitation, local programming and overall programming. In local programming, three losses of a hot end, a cold end and a heat regeneration end are expressed as ab objective function, and a non-derivative local optimal algorithm is adopted so as to achieve a high convergence speed; and in overall programming, the flow deviation of the heat end, the cold end and the heat regeneration end represents the objective function, and a heuristic global optimization algorithm is adopted so as to find a global optimal solution as much as possible. According to the method, the global optimal solution of the hot efficiency and the losses can be found through one-time calculation through overall programming outer iteration and local programming inner iteration, equation closure discussing, initial value debugging and heuristic algorithm debugging are avoided, the solving difficulty is reduced, and multi-configuration and multi-variable synchronous optimization is achieved.

Description

technical field [0001] The invention belongs to the technical field of power engineering and engineering thermophysics, and in particular relates to a modeling method of a Brayton cycle system based on nonlinear programming. Background technique [0002] Due to the strong nonlinearity brought about by the actual fluid properties and processes such as diversion and confluence, the modeling and optimization of complex Brayton cycle systems has always been one of the most difficult points in process synthesis and optimization. In the past few decades, there are usually two types of equation-based solving methods, namely sequential algorithm and simultaneous equation method. system of linear equations. [0003] A relatively mature theory has been born based on the sequential algorithm, and common commercial software such as ASPEN and EBSILON mostly adopt this method. However, the sequential algorithm needs to frequently debug unknown guess values, and set the constraints of di...

AI Technical Summary

Problems solved by technology

The main problem it faces is consistent with the sequential algorithm: the equation system must be closed before it can be solved. If the underdetermined equation is solved, the zero equation needs to be input to start the solver; if the equation is overdetermined, it cannot be solved and can only be solved based on the minimum Residuals estimated by the square method

In addition, this method cannot deal with the problem of fuzzy constraints

[0005] However, when using the method of solving nonlinear equations to model and calculate complex Brayton cycle systems, the closure of the equations must first be considered, but the existence and uniqueness of the real number solutions of the nonlinear equations are not related to the closure of the equations. There is no direct relationship, which is why it is difficult to identify multiple solutions by traditional nonlinear equation solution methods

In addition, after a series of debugging to obtain a set of steady-state solutions, it is necessary to ensure the stability of the equation system solution algorithm for subsequent optimization; and strong nonlinear problems are very sensitive to changes in initial values ​​and parameters. When the system is complex Time stability is often difficult to guarantee

Finally, if heuristic optimization is used to avoid local optima, it also faces the problem of computational inefficiency

[0006] In summary, the complex Brayton cycle system—especially one of the research hotspots in recent years: the modeling and optimization technology of the supercritical carbon dioxide Brayton cycle system, faces such problems in system modeling, solution and optimization: strong non- Linearity makes it necessary to discuss the closure of the equation system during the solution process, and repeatedly debug the algorithm to ensure its stability. In addition, it will take a lot of time in the heuristic optimization process

Images