The increasing integration of renewable energy sources (RES) and electric vehicles (EVs) into distribution networks presents operational challenges, including voltage instability, power quality issues, and heightened power flow volatility. This paper proposes a bi-level optimization framework for multi-objective energy management in active distribution networks (ADNs). The upper level employs a Linear Programming (LP) model to optimally schedule Demand Response (DR) programs, managing flexible electrical loads and EV charging stations. The lower level utilizes a Particle Swarm Optimization (PSO) algorithm to determine the optimal setpoints for Soft Open Points (SOPs) and Smart Transformers (STs), enabling precise control over active and reactive power flows. The framework minimizes total operational cost, enhances voltage stability by maximizing the minimum Voltage Stability Index (VSI), minimizes the Average Voltage Deviation (AVD), and mitigates line congestion. A fuzzy-based membership function approach, combined with a minimum Euclidean distance criterion from the ideal point, is adopted to scalarize the multi-objective problem and obtain a compromised Pareto-optimal solution. The proposed algorithm is validated on a modified IEEE 69-bus radial distribution system, integrated with photovoltaic (PV) units, wind turbines, residential/public EV charging stations, and an SOP. Simulation results across six operational scenarios demonstrate that the proposed multi-objective strategy achieves a 4.82% reduction in daily operational costs and a 15.09% improvement in the voltage stability index compared to the base case, while reducing the average voltage deviation and maximum line loading by 9.80% and 0.97%, respectively. The results validate the bi-level framework in achieving a techno-economic trade-off, optimizing the modern, renewables-rich ADNs.