Abstract
In this paper, a special method based on the neural network is presented, which is conveniently used to precompute the steps of numerical integration. This method approximates the behaviour of the numerical integrator with respect to the local truncation error. In other words, it allows the precomputation of the individual steps in such a way that they do not need to be estimated by an algorithm but can be directly estimated by a neural network. Experimental tests were performed on a series of electrical circuits with different component parameters. The method was tested for two integration methods implemented in the simulation program SPICE (Trapez and Gear). For each type of circuit, a custom network was trained. Experimental simulations showed that for well-defined problems with a sufficiently trained network, the method allows in most cases reducing the total number of iteration steps performed by the algorithm during the simulation computation. Applications of this method, drawbacks, and possible further optimizations are also discussed.
Highlights
There is currently a huge expansion in the development of artificial intelligence (AI), neural network (NN), and machine learning (ML)
We present a method for improving the computation performance of Transient Analysis of Simulation Program with Integrated Circuit Emphasis (SPICE) simulators [5] based on Numerical Integration (NI) step estimation with the utilization of NN
We presented a NI step prediction mechanism NNSP based on NN
Summary
There is currently a huge expansion in the development of artificial intelligence (AI), neural network (NN), and machine learning (ML). We present a method for improving the computation performance of Transient Analysis of Simulation Program with Integrated Circuit Emphasis (SPICE) simulators [5] based on Numerical Integration (NI) step estimation with the utilization of NN. The ultimate goal of our proposed method is to increase speed of computation by lowering number of iteration step that diverged altogether with preserved accuracy of the original solution. The proposed solution was developed exclusively for use in SPICE-like electrical circuit simulation. It has been experimentally tested by implementing it in the circuit simulation program NgSpice [12], which is a freely available open-source alternative to SPICE simulator. It is possible to reuse our proposed method for other types of simulations with only a slight modification
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call