一种公园专用环境智能监测巡检小车 Dedicated Intelligent Environmental Monitoring and Inspection Vehicle for Parks
- Led development of a dynamic forecasting engine coupling bidirectional LSTM with multi-head attention for high-frequency park temperature, humidity, light, and rainfall analysis. Optimized controller iteration (100 Hz) held prediction error at 1.8–4.2% and cut seasonal error from 31% to 6.5%, supporting irrigation, shading, and severe-weather early warning.
- Built cross-network cooperative control and active safety for the inspection rover: a PyQt5 host with encrypted cloud links and 3 s auto-reconnect command forwarding for stable remote operation outdoors. Integrated pedestrian detection (99.3% accuracy) and emergency braking for 0.5 s response within 1.5 m of obstacles, filling perception and safety gaps for dedicated park inspection platforms.
This utility model discloses an intelligent park-environment inspection rover built around a central controller integrating environmental sensing, pedestrian perception, motion/actuation, and status/interaction with a bidirectional LSTM forecasting engine. It collects temperature, humidity, light, and rainfall in real time with high-accuracy environmental prediction; supports pedestrian-aware avoidance, emergency braking within 1.5 m of obstacles, and audio-visual alarms. A graphical host plus PC, mobile, and web clients enable remote operation over encrypted cloud links with automatic model fine-tuning—addressing slow, low-coverage manual patrols with low-cost smart inspection, early warning, and multi-terminal coordination.