Speech brain-computer interface (speech BCI) aims to reconstruct speech from recorded brain signals. Real-time speech BCI relies on speech detection, which is greatly impacted by the selection of speech-related neural frequency features. However, most studies did not investigate this aspect when designing speech detectors. In this study, both electrocorticography (ECoG) dataset and stereo-electroencephalography (sEEG) dataset were utilized to investigate the impact of brain signal type on the contribution of frequency bands to speech detection. We calculated the mutual information (MI) between neural frequency bands and the audio envelope and found that the distributions of frequency bands varied between the two types of brain signals. Specifically, the 40-60Hz of ECoG signal and 0-20Hz of sEEG signal got the highest MI values. To address this, we propose a two-module detector that combines convolutional neural networks and long short-term memory (CNN-LSTM) for feature extraction and speech prediction. Our detector outperformed three commonly used detectors, including Linear discriminant analysis (LDA), Support Vector Machine (SVM), and LSTM. Notably, a high correlation was found between CNN output and the frequency bands, and high MI values were observed in both types of brain signals. These findings confirm the interpretability and generalizability of our proposed speech detector.

In this paper, a long outstanding problem, robust impact control and force regulation with guaranteed stability is solved by adding a novel positive acceleration, feedback to the feedback loop in the direction of impact and force control for robot manipulators working in an unknown environment. A quick and stable transition can be accomplished with a nonzero impact velocity, and the transient force response can be controlled with minimum impact forces and minimum bouncing. In addition, a switching control strategy is designed to guarantee the stability of the impact control. Unexpected bouncing can be eliminated after finite switches. The output force and motion can be regulated simultaneously with high accuracy after contact is established. Rigorous stability analysis based on the Liapunov-like method is presented for the proposed control scheme. The control scheme was implemented and tested on a 6 DOF PUMA 560 robot arm. The experimental comparisons with other force control strategies were given to demonstrate the advantage of the proposed method. The result obtained in this paper can be applied to any system requiring impact control and force regulation.

A new sensor-referenced control method for robot impact control and force regulation with environmental uncertainties is developed in this paper. The robot dynamic model is feedback linearized and decoupled for the free motion mode, the phase transition, mode as well as the constrained motion mode. After introducing an impact model, together with the proposed positive acceleration feedback control scheme, a stable contact can be made with a nonzero impact velocity. During the impact, large impact forces and bouncing can be avoided. Moreover, the output force can be regulated after contact is established. The scheme was implemented and tested on a 6 DOF PUMA 560 robot arm. The comparison with other control methods for robot phase transition is presented. The experimental results clearly demonstrate the advantages of the proposed method.