This paper presents a continuous-time dynamic system optimum model for Link-Based Dynamic Network Loading, which aims to determine the minimum total travel time. To account for lane-changing effects and spillbacks, a link-based double queue model is utilized and extended into the Lane-Changing double queue model by incorporating the concept of lane-changing intensity. To address the time-varying nonlinear component resulting from lane-changing intensity, a solution approach combining a heuristic genetic algorithm and a linear programming solver is proposed. Its objective is to find an optimal solution by approximating the actual values of lane-changing intensity. The effectiveness of the proposed model and solution procedure is evaluated using a network with multiple OD pairs in numerical experiments. The results provide empirical evidence of the model’s ability to capture lane-changing effects.