Energy-efficient Control Of High-speed Trains Considering Punctuality Constraint

With the development of China's high-speed railway and its operating kilometrage increasing rapidly, requirements of the energy conservation, environmental protection and improvement of system efficiency in transportation, have become more and more significant. The important implements include lightweighting of train body, train head design for resistance reduction, and optimization of onboard traction drive system. This dissertation aims to take researches on the optimization of control strategies for energy-efficent operation of high-speed train, ultilizing the spare time fully in each specified interval with safety and punctuality constraint. This subject is of great importance for energy-efficient control of high-speed train embracing both theorical and practical apsects.This dissertation firstly introduced main parameters and characteristics of high-speed train which affect train's operating process, and an energy-efficient nonlinear optimal control model of high-speed train with continual control is proposed based on some model assumptions and rationality analysis. The electric regenerative braking and pneumatic braking are separated from integrated braking characteristics for the first time to describe the electric-pneumatic braking procedure more exactly. The efficiencies of onboard traction drive system and energy utilization ratio of electric regenerative braking are introduced into energy consumption function simultaneously to calculate electric energy consumption more accurately.Based on the optimal control theory, we introduced Hamilton function to dispose the integral term of the target optimization function and the equality constraint of the state equations, and complementary slackness parameter to dispose the state inequality constaint wheneven the train speed should be no higher than speed limit for operation safety. Then the Hamilton canonical equations were obtained. Applying the Pontryagin maximum principle, we got seven optimal control modes which include full power (FP),partial power (PP), coasting (C), partial electrical braking (PEB), full electrical braking(FEB), partial integrated braking (PIB) and full integrated braking (FIB). With the help of the Hamilton canonical equations, three algebra equations of the singular control modes corresponding to different holding speeds, and four differential equations of the regular control modes for the adjoint variables were derived, which provides theorical base for numerical solution of energy-efficient control of a train.To establish the number of the optimal switching rules among the ninety-eight types and the corresponding optimal switching time between control modes, can deflate solution space and save calculation resources, which is the core of the optimal energy-efficient control algorithm. We introduced two new concepts for more convenient analysis, which are optimal switching set and optimal switching direction respectively,and classified the switching strategies according to whether the train speed touches speed limit or not. During the detailed analysis, two characteristics of the adjoint variable, which include the continuity in case of the train speed never touching the speed limit and positive jump in case of the train speed touching the speed limit. Forty-three types of the optimal switching strategies were obtained, including thirteen types of optimal switches in case of train speed trajectory below the speed limits, thirty types in case of train speed trajectory touching the speed limits, and the corresponding optimal switching times were explicitly given.Necessary conditions were analyzed via variation method for two new local optimal linkages of the neighbouring parital electrical braking sections interrupted by single steep gradient and nonsteep gradient respectively, applying full electrical braking in case of the former and coasting in case of the latter. In each case, the local optimal target function was obtained via subsection integration of the train speed among neighbouring speed-holding sections. Then two key equations were derived by introducing Lagrange multipliers to solve the extreme conditions for the speed at the start of the interrupting gradient and one at the end of the interrupting gradient. On the basis, we proved the existence of the local optimal linkage, and sufficient condition of the uniqueness, and we also analyzed the uniquess of the local optimal linkage from the geometrical view. Researches on the new two local optimal linkage above are supposed to the supplement of the optimal control theory for energy-efficient train control problem.With the help of the analysis procedure, a global optimization algorithm of the optimal energy-efficient control model based on local optimal linkages was proposed,which can solve the forced interrupt problem when the interval exists neutral zones.Moreover, the proposed optimization algorithm was expanded into engineering application by offering two modified methods for three nonsingular operational regimes—full power, full electrical brake and full integrated brake. A virtual interval comprised of complex railway profile was adopted to taking researches on the factors of the energy consumption of a train, which verified the proposed model and theoretical generality of the proposed algorithm. To verify the model and algorithm in the engineering application,we chose drivers' manipulating data of a corridor from Guiyangbei station to Huaihuanan station of Shanghai-kunming passenger high-speed railway in the upward direction, and another corridor from Douzhuang station to Nanjiao station of Handan-Changzhi heavy haul railway in downward direction, to take comparison with the simulation data obtained from the simulation platform. We can conclude from the two comparisons above that the optimal energy-efficient control algorithm is valid of the optimization of high-speed train considering safety and punctuality constraint, and heavy haul train with some restrictions.