CN-122004894-A - Self-powered wearable brain electricity emotion recognizer and preparation and recognition method thereof

Abstract

The invention provides a self-powered wearable electroencephalogram emotion recognizer and a preparation and recognition method thereof, which belong to the fields of wearable electronic equipment, emotion calculation and man-machine interaction, wherein the recognizer integrates a self-powered module, a sensing module and a control and processing module, the self-powered module comprises a flexible organic photovoltaic device and is used for converting ambient light energy into electric energy, the sensing module comprises a multi-channel electroencephalogram sensor array and is used for synchronously collecting EEG signals, the control and processing module is integrated through a flexible printed circuit board and is used for data processing and wireless transmission, the method comprises the steps of carrying out spatial oversampling on electroencephalogram signals at a single position through the multi-channel EEG sensor array and evaluating the quality of each channel signal in real time, autonomously selecting optimal EEG signals to actively alleviate the interference of motion artifacts, and then processing the optimal EEG signals through a deep learning process to extract emotion characteristics based on the EEG signals and generate emotion classification results, so that stable, autonomous and continuous emotion monitoring without intervention is realized.

Inventors

• ZHENG DING
• ZHANG XIU
• ZHANG XIYUAN
• YU JUNSHENG

Assignees

• 电子科技大学

Dates

Publication Date

20260512

Application Date

20260212

Claims (7)

1. 1. The self-powered wearable brain-electricity emotion recognizer is characterized by comprising a self-powered module, a sensing module and a control and processing module, wherein the self-powered module comprises a flexible organic photovoltaic device used for converting ambient light energy into electric energy and supplying power for the recognizer, the sensing module comprises a multi-channel EEG sensor array used for synchronously collecting EEG signals, the control and processing module is integrated through a flexible printed circuit board and used for data processing and wireless transmission, and the self-powered module, the sensing module and the control and processing module are packaged in a flexible biocompatible elastomer.
2. 2. The self-powered wearable electroencephalogram emotion recognizer according to claim 1, wherein the flexible organic photovoltaic device is of an inverted structure, and the flexible organic photovoltaic device comprises a flexible transparent cathode ITO-PET, a zinc oxide transmission layer, a PM6:Y6 bulk heterojunction active layer, a molybdenum trioxide hole transmission layer and a silver anode in sequence from bottom to top.
3. 3. A self-powered wearable electroencephalogram emotion recognizer according to claim 1 and wherein said multichannel EEG sensor array comprises an array of 8 dry electrodes configured for signal acquisition on the forehead for spatial oversampling of a single location.
4. 4. A self-powered wearable electroencephalogram emotion recognizer according to claim 1 wherein said control and processing module comprises a microcontroller unit, a 24-bit analog-to-digital converter and a wireless transmission module integrated on a 4-layer flexible printed circuit board.
5. 5. A method for manufacturing a self-powered wearable electroencephalogram emotion recognizer according to claims 1-4, comprising the steps of: a1, cleaning an ITO-PET substrate, sequentially performing ultrasonic treatment in diluted detergent, acetone, deionized water and isopropanol, and then performing ultraviolet treatment for 20 minutes; a2, spin-coating zinc oxide precursor solution on the ITO-PET substrate to form a zinc oxide transmission layer for electron transmission, and annealing for 15 minutes at 150 ℃; A3, depositing an active layer solution on the zinc oxide transmission layer by adopting a spin coating process in a glove box filled with nitrogen to form a bulk heterojunction active layer, and annealing for 10 minutes at 100 ℃; A4, sequentially depositing a molybdenum trioxide hole transport layer with the thickness of 15nm and a silver anode with the thickness of 100nm by thermal evaporation under the condition that the pressure of the molybdenum trioxide hole transport layer is less than 1e-5mbar, so as to finish the preparation of the flexible organic photovoltaic device; A5, integrating a control circuit assembly comprising a micro controller unit and an analog-to-digital converter on a 4-layer flexible printed circuit board fPCB; A6, performing system integration on the prepared flexible organic photovoltaic device, fPCB and the multichannel EEG sensor array, and connecting an external reference electrode through a lead wire; And A7, packaging the whole assembly except the external reference electrode by using a medical grade silica gel elastomer to form the final patch type identifier.
6. 6. A method for emotion recognition of a self-powered wearable brain-electrical emotion recognizer, applied to the recognizer of claims 1-4, comprising the steps of: B1, synchronously acquiring multichannel EEG signals through the identifier; B2 inputting representations of the plurality of EEG signals into a VGG16 model, said model being trained to evaluate signal quality, thereby autonomously selecting an optimal one-dimensional EEG time series signal; B3, converting the optimal one-dimensional EEG time sequence signal into a two-dimensional time-frequency spectrogram through continuous wavelet transformation, and inputting the time-frequency spectrogram into a visual transducer model to extract and classify high-dimensional emotion characteristics.
7. 7. The method of claim 6, wherein B3 said visual transducer model captures long range dependencies and global context information in the time-frequency spectrogram via its internal multi-headed self-attention mechanism to identify emotional states.

Description

Self-powered wearable brain electricity emotion recognizer and preparation and recognition method thereof Technical Field The invention belongs to the fields of multi-mode information processing, artificial intelligence and emotion calculation, and particularly relates to a self-powered wearable brain electricity emotion recognizer and a preparation and recognition method thereof. Background Emotion computing is a interdisciplinary discipline aimed at recognizing and interpreting human emotion, and has made great progress in recent years with the development of sensor design, circuit integration, and algorithm optimization. By capturing physiological signals such as speech, facial expressions, and brain waves (EEG), emotion calculation can quantitatively and qualitatively acquire a wide range of human emotion information. Compared with other modes, emotion recognition based on physiological signals such as EEG is more objective and is not easy to control subjective consciousness. However, existing emotion recognition techniques still face serious technical challenges. First, most emotion recognition systems are limited in their power supply and portability. The existing system generally depends on the traditional battery or wired connection to supply power, which not only limits the application of the device in a long-term non-invasive monitoring scene, but also increases the volume and weight of the device, and prevents the portable use of the user. The need to frequently charge or replace batteries makes continuous and uninterrupted emotion monitoring difficult to achieve. Second, EEG has many limitations as a tool for assessing brain state in traditional applications. Clinical grade EEG equipment is often bulky, expensive, requires the use of conductive gel, and requires the test subject to remain stationary, which greatly limits its usefulness in everyday life. Furthermore, the operation of the device often requires the assistance of a professional, with operator dependencies, resulting in possible differences in the results between different operator measurements. More critical is that the wearable EEG sensor is extremely prone to losing the target signal during the user's movements. For sensors placed on the forehead, the facial expression, speaking and head movements of the user can introduce serious motion artifacts and poor electrode contact, can seriously affect EEG signal perception, and is one of the core problems to be solved by the wearable EEG system. Thus, users are often required to be stationary to ensure data quality, which is contrary to the actual application scenario. While wearable EEG techniques have made some progress in overcoming some of the problems described above, the prior art fails to address these interrelated challenges in that wearable EEG devices are greatly affected by motion artifacts while device portability using large batteries or wired connections is reduced. Disclosure of Invention In order to solve the problems in the background art, the invention provides a self-powered wearable electroencephalogram emotion recognizer and a preparation and recognition method thereof, which are used for solving the problems that a traditional battery or wired connection limits the continuous working time of equipment, increases the volume and the weight, simultaneously reduces portability, motion artifacts seriously interfere electroencephalogram (EEG) signals, body motion and facial expression cause data quality reduction and even signal loss, limit the effectiveness of measurement, and the traditional EEG system is complicated to operate and depends on professionals, needs to use conductive gel and additional equipment, and is not suitable for daily deployment. In order to achieve the above purpose, the present invention provides the following technical solutions: A self-powered wearable brain-electric emotion recognizer comprises a self-powered module, a sensing module and a control and processing module, wherein the self-powered module comprises a flexible organic photovoltaic device used for converting ambient light energy into electric energy and supplying power for the recognizer, the sensing module comprises a multichannel EEG sensor array used for synchronously collecting EEG signals, the control and processing module is integrated through a flexible printed circuit board and used for data processing and wireless transmission, and the self-powered module, the sensing module and the control and processing module are packaged in a flexible biocompatible elastomer. Preferably, the flexible organic photovoltaic device adopts an inverted structure, and the flexible organic photovoltaic device comprises a flexible transparent cathode ITO-PET, a zinc oxide transmission layer, a PM6:Y6 bulk heterojunction active layer, a molybdenum trioxide hole transmission layer and a silver anode from bottom to top. Preferably, the multi-channel EEG sensor array comprises an array of 8 dry ele