Design and implementation of intelligent voice IoT control system based on dual core architecture

This design proposes an intelligent IoT voice control system based on a dual core architecture, which achieves high-performance computing and real-time communication through the collaborative division of labor between ESP32S3 and ESP-01S microcontrollers. The system adopts a separated architecture design, and the voice control module is built on the ESP32S3 with the Xiaozhi ESP32 open source framework, combined with the ASR-PRO core board to achieve low-power voice wake-up function. It supports user-defined wake-up words and completes semantic parsing and intent recognition of online voice instructions through the LLM large model and MultiNet command word recognition model. The communication control module is based on ESP-01S to build a multi protocol IoT hub with MQTT as the core, integrating MQTT based multi protocol stacks to achieve bidirectional interaction between the device layer and the cloud. Real time sensor data is collected through UART interface and transmitted to the cloud platform via Wi-Fi network via MQTT protocol, while parsing and executing control instructions from mobile terminals. The system realizes a complete link from voice wake-up, instruction recognition to device control, ensuring real-time response performance while having good scalability.