In wireless or sensor networks, due to limited resources, dynamic network topology, unreliable communication links, and malicious intrusion, it is still difficult to find secure and energy efficient routing schemes. In this study, a graph based modelling and multiobjective optimization model for secure routing in wireless sensor networks is proposed. The sensor nodes and communication links are modelled as a weighted graph which includes link reliability, node trust, link energy consumption, transmission delay and security risks. A composite routing cost function dynamically assesses possible paths, and discards nodes that are malicious, unreliable or have little energy. The proposed framework provides a route selection mechanism that can be adapted to changing scenarios on network and security. Experimental examination indicates that the packet delivery, attack avoidance, energy efficiency, communication delay and network robustness are superior to the conventional shortest-path, energy-aware, and trust-based routing methods. Dynamic trust evaluation adds more to the anti-blackhole and anti-selective-forwarding protection. The system design proposed offers a scalable and resource-efficient solution to ensure reliable, adaptable, and secure communications for future wireless sensor and IoT systems.