[Purposes]With the increasing complexity of urban traffic, the energy consumption of intelligent vehicles in the multi-signal scene has been paid more and more attention. The purpose of this study is to explore an effective method to improve the energy consumption economy of intelligent vehicles in multi-signal scenarios, and focus on reducing the energy consumption generated by vehicles in urban cruising conditions, so as to optimize the energy utilization efficiency of vehicles. 【Methods】A dynamic programming-quadratic programming (DP-QP) framework combining dynamic (DP) quadratic programming (QP) is proposed, which transforms the optimal control problem of speed planning into a nonlinear optimization problem. Firstly, the passing area of the vehicle is divided by randomly scattering points, then the path of the preliminary planning is optimized by dynamic programming, and then the path is smoothed by quadratic programming, and the smoothed trajectory is taken as the initial iteration value of the quadratic optimization, so as to accelerate the convergence process of the optimal speed trajectory. In order to verify the validity of the framework, the vehicle speed planning model was built on the Matlab/Simulink platform, and the joint simulation test was carried out with Carsim and PreScan. At the same time, the irregular signal light phase and timing (SPaT) scenario was constructed. [Findings]The simulation results show that the DP-QP method has achieved remarkable results. Compared to the M-IDM algorithm, DP-QP can reduce energy consumption by 21%. At the same time, the calculation time of DP-QP is nearly 14 times faster than that of SSO algorithm. [Conclusions] To sum up, the DP-QP algorithm successfully finds the best balance point between real-time performance and energy-saving effect, and has obvious advantages in multi-signal intersection scenarios, which can provide strong support for energy consumption optimization of intelligent vehicles in urban working conditions.