Rice cultivation accounts for 65% of agricultural water consumption in China but suffers from low irrigation water use efficiency (WUE ≈ 0.80kg/m^3) and severe nitrogen and phosphorus loss, contributing up to 30% of agricultural non-point source pollution. To address these core issues, this study developed a multi-source collaborative sensing-based Intelligent Precision Paddy Irrigation Control System (IPICS). The system integrates satellite and UAV remote sensing with IoT technologies to construct a ``sky-space-ground'' three-dimensional monitoring system, couples the FAO-56 Penman-Monteith model with LSTM deep learning algorithms to establish a dynamic irrigation decision model, and integrates a self-developed solar-powered intelligent sluice gate (AG-SOLAR-M7) for precise canal water control. A two-year field validation (2023-2024) was conducted at Youyi Farm (131.8121°E, 46.7825°N), comparing the Intelligent Precision Irrigation Control System (IPICS) with conventional flooding(each 40 ha, 3 replicates). Results demonstrated that: (1) IPICS reduced irrigation by 25.3% (625±32 mm vs. 837±41 mm, p<0.01); (2) maintained stable yield (7.0±0.4 t/ha vs. 7.1±0.5 t/ha); (3) increased WUE by 31.8% (1.12±0.08 kg/m^3 vs. 0.85±0.06 kg/m^3); (4) decreased nitrogen leaching by 57.4% (18.2±3.1 kg/ha vs. 42.7±5.9 kg/ha). The ``sensing-decision-execution'' closed-loop control of IPICS significantly enhanced water and fertilizer utilization efficiency, offering a replicable technical model and practical application for digital farmland development.