An Explicit Linearized State-Space Technique for Accelerated Simulation of Electromagnetic Vibration Energy Harvesters

Abstract

Abstract-Vibration energy harvesting systems pose significant modeling and design challenges due to their mixed-technology nature, extremely low levels of available energy and disparate time scales between different parts of a complete harvester. An energy harvester is a complex system of tightly coupled components modeled in the mechanical, magnetic as well as electrical analog and digital domains. Currently available design tools are inadequate for simulating such systems due to prohibitive CPU times. This paper proposes a new technique to accelerate simulations of complete vibration energy harvesters by approximately two orders of magnitude. The proposed technique is to linearize the state equations of the system's analog components to obtain a fast estimate of the maximum step-size to guarantee the numerical stability of explicit integration based on the Adams-Bashforth formula. We show that the energy harvester's analog electronics can be efficiently and reliably simulated in this way with CPU times two orders of magnitude lower than those obtained from two state-of-the art tools, VHDL-AMS and SystemC-A. As a case study, a practical, complex microgenerator with magnetic tuning and two types of power processing circuits have been simulated using the proposed technique and verified experimentally.

Generation of new power processing structures exploiting genetic programming

“…Since power processing circuits comprise diodes, which are nonlinear components, it is necessary to linearize their models to create linearized state-space equations. Following the method proposed in [6] The integration scheme requires five types of equations for each circuit. These equations are dynamically generated by the program and then evaluated using the integration scheme.…”

“…In order to speed up the fitness evaluation, SPICE simulations are replaced by a explicit integration method based on the Adams-Bashforth formula. This technique lies on the linearization of the state equations of the analog components of the circuit and allows a bigger maximum step-size preserving the numerical stability of the integration algorithm [6]. The rest of the paper is organized as follows: Section II describes the genetic algorithm employed in this work.…”

Fast Design Space Exploration of Vibration-Based Energy Harvesting Wireless Sensors

“…The Authors are with the Faculty of Physical Sciences and Engineering, University of Southampton, Southampton, SO17 1BJ, UK, (phone: +44 2380593520; fax: +44 2380592901; email: {tjk,lw04r,gvm,bmah,ma08r}@ecs.soton.ac.uk) technique which can reduce the CPU time of one simulation by two orders of magnitude [7], it is still not feasible to optimize the energy performance of a complete wireless sensor node where many thousands of simulations are required. Also, although classical multi-variable optimization methods based on gradient searches and various heuristic algorithms, such as genetic optimization or simulated annealing, allow improvements to an initial design, they do not lend themselves easily to fast design space exploration and investigation of possible trade-offs.…”

“…Full simulations are carried out for each of the design points to obtain the performance indicator values from which the RSM coefficients are obtained. Although the number of the design points is moderate, from 21 to 78 in our test scenarios, the use of the fast linearized state-space technique [7] allows the RSMs to be built in less than 5 hours. Once the RSMs are built, exploration of the design space and investigation of trade-offs between the design parameters and performance indicators is instantaneous.…”

DoE-based Performance Optimization of Energy Management in Sensor Nodes Powered by Tunable Energy-Harvesters

“…However, these solutions have not fully solved the main challenge of the traditional analogue simulation approach based on Newton-Raphson iterations which are the main cause of the long CPU times. We have developed a linearized statespace technique which can reduce the CPU time of one simulation by two orders of magnitude [4], but, as many thousands of simulations are required in traditional, simulationbased optimisation, it is still not possible to design efficiently a complete wireless sensor node. Classical multi-variable optimization methods are based on gradient searches and various heuristic algorithms, such as genetic optimization or simulated annealing.…”