A novel model order reduction technique based on artificial intelligence

Computer aided design (CAD) plays a vital role in modern VLSI design. Electronic simulation is widely used to verify the design and test the behavior of the circuit before fabrication. One of the major research areas in CAD is circuit simulation. Circuit simulation is to use mathematical models to predict the behavior of an electronic circuit. A circuit is usually represented by a set of partial differential equations (PDEs) or ordinary differential equations (ODEs). So, the circuit simulation actually involves solving large-scale ODEs which sometimes takes several days or even weeks. Therefore, fast and accurate circuit simulation algorithms are needed to accelerate the simulation cycle. One way to speed up the simulation is to approximate the original system with an appropriately simplified system which captures the main properties of the original one. This method is called model order reduction (MOR), which reduces the complexity of the original large system and generates a reduced-order model (ROM) to represent the original one. There are many existing MOR methods, but there is no method that gives the best results for all of the systems. So, each system uses the best method according to its application. So, there is still a need for novel MOR techniques. This paper presents a novel MOR technique based on artificial intelligence.