To provide end-to-end, deterministic bandwidth, delay, packet loss, delay jitter and other network QoS guarantees for diverse electrical services is one of the critical tasks in communication network underlaying ubiquitous electrical Internet of things. In this paper, a differentiated QoS routing strategy based on cooperative intelligence and subgradient optimization algorithm is proposed to address the issues of route selection under multiple constraints, such as slow convergence, easy to fall into sub-optimal solution, lack of feasibility test of solution and difficult to prove optimality. Specifically, the sub-gradient method can dynamically adjust the weighting coefficients associated with multiple QoS constraints, so as to perform feasibility check of solutions during the iteration. In addition, we put forward an improved coordination mechanism for basic ant colony algorithm, which can be used to improve the discrimination of different links and thus achieve fast convergence and avoidance of entering the local sub-optimal solution. The simulation results show that our proposed algorithm can converge to the optimal routing path faster than the existed heuristic algorithms. Besides, the proposed algorithm provides the verification of solution feasibility and optimality.#$NLKeywords: swarm intelligence optimization; subgradient optimization; ubiquitous electrical Internet of things; dynamic routing strategy