Putra Wisnu Agung Sucipto, Rahmat Zidan, Sujud Purnomo Subiyantoro, Annisa Firasanti, Eki Ahmad Zaki Hamidi
This research develops a distributed state estimator framework that integrates stochastic triggers and dissipative filters to improve communication efficiency and state estimation accuracy in dynamic environments. The proposed method using a probability-based stochastic trigger policy to propagate the current state. If no state is transmitted, the swarm is required to estimate remotely using a dissipative filter using a dynamic stochastic error criterion. This framework was tested on a swarm of omnidirectional robots using a webot simulator to evaluate the communication efficiency between stochastic and time-triggered transmissions. The estimation accuracy of the dissipative filter was also evaluated against the Kalman filter, particle filter, GHK filter, and Bayes filter. Experimental results show that transmission reduction can occur when using this stochastic trigger by 25%, which is smaller than that of the time-triggered filter. The dissipative filter outperforms the standard filter in accuracy with an MEA of 0.3610 and an RMS in the range of 0.314-0.497. The estimation-to-transmission ratio increases from 1.24 under time-triggered, to 2.13 when using stochastic-triggered. These results demonstrate that this framework effectively reduces communication overhead while maintaining estimation accuracy. © 2025 IEEE.
State University of Malang, Department of Electrical Engineering and Informatics, Malang, Indonesia; University of Islamic 45, Department of Electrical Engineering, Bekasi, Indonesia; Uin Sunan Gunung Djati, Department of Electrical Engineering, Bandung, Indonesia
