Parameter Optimization Based On Multi-domain Model Analysis

Along with advance in science and technology, modern products are increasingly complex. Modern electromechanical products generally are composed of subsystems from multiple engineering domains such as mechanic, electric, hydraulic, and control system components, and traditional single-domain simulation tool can not meet the needs of its design and analysis. In recent years the Modelica language has been fully developed in the analysis and design of complex electromechanical products. So parameter optimization based on multi-domain model analysis is the trend of multi-domain modeling and simulation and owns fine technical application foreground.This paper is based on the development of platform MWorks, which is a multi-domain physical modeling and simulation system developed by CAD center of HUST independently, the problem of parameter optimization based on model analysis in multi-domain is studied, and it is the first domestic material discussing the simulation optimization based on Modelica language and multi-domain modeling and simulation platform. Parameter optimization is studied from the aspects of multi-domain optimization modeling, optimization flow design, optimization strategy and so on.Multi-domain optimization modeling is the precondition of parameter optimization. The mixed expression of simulation and optimization model and the value type of the state variable and the design function (include constraint function and object function) are presented, during which the method of computing the value of the criteria of the kind RiseTime, SettlingTime and Overshoot is introduced in detail. Multi-case optimization modeling, which supports several plans optimized simultaneously and heuristic optimization modeling method based on complier technology are both studied and achieved.Optimization solving strategy is the key points of parameter optimization. The design of optimization flow is studied and realized firstly, transplant and encapsulation of many solving algorithms are accomplished, and finally the solving strategy self-adapted option aim at specific optimization algorithm is proposed based on optimization model attribute and optimization algorithm characteristic analyzing.On the basis of the above achievement, parameter optimization based on multi-domain model simulation has been carried out in the module of MWorks /Optimization. At the end of this paper, a typical example of parameter optimization is given, and the validity of optimization algorithm is proved.