Moving Horizon Estimation on a Chip
Final rept. 1 Jun 2012-31 May 2014
NANYANG TECHNOLOGICAL UNIV (SINGAPORE)
Pagination or Media Count:
This work investigated the feasibility of solving Moving Horizon Estimation MHE problems in hardware, e.g. Field Programmable Gate Arrays FPGA, in contrast to using software. With the hardware approach we could build customized computational machines for specific algorithms and hence gain computational and power efficiency, which is crucial for embedded real-time applications where computational resources are limited. A hardware approach enables one to build certifiable components and is also less susceptible to software virus attack. Using high level synthesis tools instead of low level hardware description languages, we demonstrated that it is feasible to routinely deploy customized, sophisticated computational algorithms such as MHE, on reconfigurable hardware for real-time embedded applications.
- Operations Research
- Navigation and Guidance